SystemPage now only shows operational content: MITRE Sync, Intel Scan,
ATT&CK Evaluations, Scheduled Jobs, and Template Management.
Settings gets two new admin-only tabs:
- "SSO / Azure AD": full SAML 2.0 wizard (5-step setup for Azure AD)
- "System": System Status + Version Info + Configuration Export/Import
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- sso_service: fix process_callback for Azure AD claim URIs (email, role)
- Default role_attr to full Azure role claim URI
- Fallback email resolution via Azure email claim URI + NameID
- Username defaults to full email (prevents collision with local accounts)
- User lookup also tries email field for existing local accounts
- Logs warning when unknown role received from IdP
- frontend/api/sso.ts: new API module with getSsoStatus, getSsoConfig, updateSsoConfig
- LoginPage: redesigned for SSO-first flow
- Shows Azure SSO button as primary when SSO enabled+configured
- Local login collapsed under "Emergency admin access" section
- Falls back to normal local login form when SSO is disabled
- SystemPage: new SsoConfigSection component (guided 5-step wizard)
- Step 1: Copy SP Entity ID and ACS URL for IT team + metadata XML download
- Step 2: Azure App Roles reference table (6 roles with exact values)
- Step 3: Tenant ID field auto-fills idp_entity_id and idp_sso_url
- Step 4: X.509 certificate paste field
- Step 5: Attribute mapping pre-filled with Azure AD claim URIs
- Enable/disable toggle + save
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Capture Step.Description (HTML stripped), step name/number, substep ref,
criteria, and data sources from MITRE ATT&CK Evaluations API
- _aggregate_by_technique() now accumulates ALL occurrences per technique
(multiple substep refs, criteria, step contexts) instead of keeping only
the best-scoring one
- New helper functions _build_procedure_text(), _build_description(),
_build_red_summary() generate rich narratives from accumulated occurrences
- New re_enrich_evaluation_round() service function + POST endpoint
/system/attck-evaluations/re-enrich to update already-imported tests
without changing detection results or validation state
- Frontend: Re-enrich button per imported round + result banner in SystemPage
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The /api/results/ endpoint returns a LIST: [{name: crowdstrike, adversaries: [...]}]
Previous code called data.get() on the list → AttributeError crash on every import.
Fix: detect list vs dict response, extract the crowdstrike vendor entry first,
then get its adversaries list. Keeps legacy dict fallback just in case.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Fallback names now use hyphens matching live API (carbanak-fin7, wizard-spider-sandworm)
- Add APT3 (R1) and Enterprise 2025/er7 (R7) to fallback - verified from live API
- Remove OilRig (R6) from fallback - CrowdStrike did not participate in Round 6
- Orange fallback banner only shows when NO rounds are available at all
- Soft gray note when rounds are loaded but API had transient error
- Check-new and import errors: detect 502/Cloudflare messages and show user-friendly text
instead of raw Cloudflare HTML error messages
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add browser User-Agent and Referer headers to all evals.mitre.org requests
- fetch_rounds_with_status() returns api_reachable flag + rounds list
- Fallback to 5 known public CrowdStrike rounds (APT29/R2 through OilRig/R6)
when live API is blocked, so UI always shows something actionable
- Router returns {rounds, api_reachable, api_error} instead of plain array
- Frontend shows orange warning banner when using fallback data
- Remove 502 HTTPException - rounds are always returned (live or fallback)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Replaces the minimal bottom legend with a full coverage legend panel
placed above the filters. Each status shows a cell mock matching the
exact colors used in the matrix, a color-coded label, and a short
description of what it means. Includes review_required with its
orange alert-triangle badge. Removes the old minimal bottom legend.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
New drag-and-drop section at the bottom of the Import RT page so operators
can convert screenshots to base64 without leaving the page. Includes
thumbnail preview, copy-base64 and copy-JSON-snippet buttons with
2s feedback, per-image delete and clear-all.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Each technique in the RT import JSON now requires at least one evidence
image (PNG/JPG/GIF/WebP/BMP, max 10 MB decoded) embedded as base64.
Backend:
- RTEvidenceEntry model: filename, data (base64), caption (optional)
- RTTechniqueEntry.evidence is now required
- Pre-validation raises 422 if any technique is missing evidence
- After test creation, images are decoded and stored in MinIO as
Evidence records (team=red) linked to the test
Frontend:
- RTEvidenceEntry type added to api/tests.ts
- parseJson() validates evidence presence and structure per technique
- Preview table shows base64 thumbnails (up to 3 + overflow count)
- Format reference updated: evidence fields moved to Required section
- Import result shows total evidence images attached
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The previous name implied data from a dedicated threat intelligence team.
The feature actually monitors public RSS feeds and security blogs for
ATT&CK technique mentions, so Security Feed Monitor is more accurate.
Updated description and all references across SystemPage and ReviewQueuePage.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Overall Security Score renamed to Overall Programme Score. Descriptions across
Executive Dashboard and Dashboard page now clarify scores reflect Red/Blue Team
exercise maturity and coverage breadth, not the organisation real-world security
state, to avoid overstating what ATT&CK simulation tests can guarantee.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
build_threat_actor_layer was adding ALL techniques to the layer —
actor techniques with their real score and non-actor techniques with
score=0/enabled=False. This caused every tactic column to appear in
the matrix even when the actor has no techniques for that tactic.
Now only actor techniques are included. The frontend already filters
visible tactics to those with data, so empty tactic columns disappear
automatically.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
client.ts: when FastAPI detail is an object, extract .message for the error
string and preserve the full detail on enhancedError.detail so consumers
can inspect structured error payloads (e.g. 409 start_date_in_future).
CampaignDetailPage: use enhancedErr.status (not response.status) and
enhancedErr.detail (not response.data.detail) to detect 409 and show
the confirmation modal instead of the toast.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
FastAPI wraps error bodies as {detail: string | object}, not at the
top level. Was reading data.message instead of data.detail.message,
causing [object Object] in the toast for all non-409 errors.
Now correctly extracts:
- 409 with object detail -> start_date warning modal
- Other errors with string detail -> readable toast message
- Other errors with object detail -> detail.message in toast
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Adds a Generate Campaign button (purple, visible to leads/admin) in the
threat actor header. Opens a modal with:
- Actor name shown as context
- Start date picker (required — validated: must be today or future)
- Warning message showing when tests will be queued
- Error display for API failures
- On success: redirects to the new campaign detail page
Start date is mandatory here (unlike the CampaignsPage flow where it
is optional) to enforce scheduling discipline when generating from actors.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Backend: activate endpoint returns 409 with structured warning when
start_date is in the future; accepts force=true to bypass.
test_crud_service: always excludes tests from draft campaigns with future
start_date so they do not appear in the team queue prematurely.
Frontend: catches 409 on activate and shows amber confirmation modal
with Keep scheduled / Activate now anyway options.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Backend:
- campaign_service.generate_campaign_from_threat_actor: accept optional
start_date kwarg and set it on the Campaign model
- campaigns router: new GenerateFromActorPayload schema, /from-threat-actor
endpoint now accepts optional body with start_date
Frontend:
- generateCampaignFromThreatActor API: accept optional options param
- Generate Campaign modal: date picker + warning message, same UX as the
manual create form
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- package.json: bump axios constraint from ^1.13.5 to ^1.14.0
- Dockerfile build stage: npm ci -> npm install so the semver range
in package.json is honoured at build time (npm ci uses the lockfile
exactly, bypassing the updated constraint)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Bumps minimum Axios version from 1.13.5 to 1.14.0 to remediate
CVE-2026-42043 identified by VMT / Wiz (asset: AegisTest).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
DB: migration b047 adds start_date (DateTime nullable) + index to campaigns.
Backend:
- Campaign model: start_date field
- CampaignCreate/Update schemas: accept start_date (ISO string)
- CRUD service: persist + serialize start_date in both serializers
- Activation endpoint: blocks manual activation if start_date is in the future
(campaign will auto-activate via scheduler)
- Scheduler: new hourly job _run_scheduled_campaign_activation — finds draft
campaigns with start_date <= now, activates them, creates Jira tickets,
notifies red_tech team
- Jira: campaign + test tickets now include JIRA_START_DATE_FIELD (configurable,
default customfield_10015). Campaign uses start_date if set, else created_at.
Tests inherit campaign start_date.
- config.py: JIRA_START_DATE_FIELD setting
Frontend:
- Campaign type: start_date field on Campaign + CampaignSummary
- CampaignCreatePayload: start_date optional field
- Create form: date picker with min=today, warning message explaining behavior
- Campaign detail header: start_date badge showing days remaining or started date
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Progress and Timing now share a 2-column grid at the top of the detail page.
Removed CampaignTimingPanel from the bottom Jira section.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Amber banner for DORA and ISO 27001:2022 — community-based mapping, no official CTID source.
Orange banner for ISO 42001:2023 — experimental, MITRE ATT&CK has no AI-specific techniques yet.
Each notice explains the mapping source, limitations, and what executives should consider
before using the data in formal audits or regulatory submissions.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Backend: expose description in control status response, add rich business-language
descriptions to all curated controls (ISO 27001, ISO 42001, CIS v8, DORA) explaining
requirements and ATT&CK mapping rationale. ISO 42001 includes infrastructure-mapping note.
Frontend: description field in type, info panel in ControlsTable expanded rows,
framework info banner with description and official standard link in CompliancePage.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
ISO 27001:2022: 37 Annex A controls across 4 themes (Organizational,
People, Physical, Technological) mapped to MITRE ATT&CK techniques.
ISO 42001:2023: 25 Annex A controls for AI Management Systems mapped to
relevant ATT&CK techniques covering AI supply chain, data pipeline
integrity, model serving security, and third-party AI risk.
Backend: import functions, _import_curated_framework() shared helper,
and POST /compliance/import/iso-27001 + iso-42001 endpoints.
Frontend: API client functions + import buttons in CompliancePage.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Lead who approved: Request Discussion button becomes Discussion Requested after sending.
Lead who rejected: new Change to Approved button to resolve conflict after offline discussion.
Both leads retain vote-change buttons. discussionSent state flag tracks send status.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- request-discussion endpoint: add 'admin' to allowed roles
- Return rejector_email and rejector_role in the response
- Modal success state shows contact card with username, role, email link
so the approving lead can immediately reach out to the rejecting lead
Backend: POST /tests/{id}/request-discussion
- Only callable by the lead whose vote is 'approved' in a disputed test
- Sends notification to the rejecting lead: 'Lead X confirms their
approval and wants to discuss your rejection'
- Logs the action in audit trail
Frontend:
- 'Confirm My Validation' button (amber outline) alongside 'Change to Rejected'
- Opens a modal showing:
* Explanation: both leads must agree to finalise
* Other lead's rejection reason/notes
* What happens next (stays disputed, notification sent, either can change)
- 'Send Discussion Request' → calls the new endpoint → shows success state:
'Lead username has been notified...'
- Instruction to reach out via team channels to resolve offline
Flow summary for disputed tests:
Approving lead sees 2 options:
a) 'Confirm My Validation' → modal → send request → other lead notified
b) 'Change to Rejected' → validation modal → both agree to reject → rejected
test_entity.py has its own TestState enum separate from domain/enums.py.
Only domain/enums.py was updated, causing AttributeError when SQLAlchemy
tried to map 'disputed' from DB to the test_entity.TestState class.
Also adds disputed to VALID_TRANSITIONS so the entity can transition
into and out of the disputed state.
1. New 'disputed' state — one lead approved, the other rejected:
- Both approved → validated (unchanged)
- Both rejected → rejected (unchanged)
- One approves + one rejects → disputed (new)
- DB: ALTER TYPE teststate ADD VALUE 'disputed'
- Notification sent to the approving lead explaining the conflict
with the rejection notes
2. Disputed UI in TestDetailHeader:
- Amber banner showing conflict + rejection reason from notes
- 'Change Vote to Rejected' button for the lead who approved
- Validation indicators shown for disputed state too
3. Fix timestamps on reopen (rejected → draft):
- Keep red_started_at, blue_started_at etc. as historical record
- Only clear paused_at defensively
- Timestamps naturally update when test is re-executed
4. disputed badge (amber) added to all badge color maps
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Backend (reopen_test):
- Preserve red/blue validation NOTES — teams see exactly what to fix
without losing the rejection context. Previously both notes were cleared.
- Preserve all content fields: procedure_text, tool_used, red_summary,
attack_success, blue_summary, detection_result (already the case).
- Preserve evidences (separate table, unaffected — already the case).
- Still clear: validation statuses + who/when validated (fresh re-validation
required). Phase timing reset so the new execution starts clean.
Frontend:
- Button label: 'Reopen Test' → 'Continue Test' (more accurate intent)
- Dialog title: 'Reopen Test' → 'Continue Test'
- Dialog message: replaces alarming 'workflow will be restarted / clear all'
with accurate description of what is preserved vs reset
- Toast: explains what to do next
useMemo executes its factory immediately on first render. lastActivityDate
was defined with const after the useMemo call inside the component, causing
a temporal dead zone: 'Cannot access v before initialization'.
Fix: move the function to module scope (before the component), where it
is fully initialized before any hook runs.
TestsPage 'Updated' column: compute lastActivityDate() from the most
recent available timestamp — blue_validated_at > red_validated_at >
blue_work_started_at > blue_started_at > red_started_at > created_at.
Also fixes the sort-by-updated_at case.
ValidatedTestsPage 'Validated' column: use blue_validated_at (when Blue
Lead approved) falling back to red_validated_at. Fixes both the display
and the default sort-by-validated.
updated_at column does not exist in the tests table — it was always
undefined, so formatElapsed() always returned '—'.
Replace with blue_started_at (set when Red Team submits to Blue Team),
which correctly shows how long a test has been waiting for Blue Team
evaluation. Also fixed the waiting_time sort to use the same field.
- Global document paste listener captures image/* items from clipboard
- Auto-generates filename: screenshot-YYYY-MM-DDTHH-MM-SS.png
- Brief cyan pulse animation confirms the paste was detected
- Shows image preview before uploading (max-h 192px, object-contain)
- Drop zone hint now says 'Drag & drop, browse, or Ctrl+V to paste'
- Works with any source: OS screenshot (PrintScreen/Cmd+Shift+4),
browser Inspect screenshots, any image copied to clipboard
Root cause: avg times were ~2-3 minutes (< 1h). round(0.033, 1) = 0.0
which is falsy in JS, so the frontend showed N/A instead of the value.
Fix (backend): _safe_stats() and team metrics now convert to minutes
when avg < 1 hour, adding a 'unit' field ('min' or 'hrs').
Fix (frontend): use != null instead of truthy check for avg_completion_hours,
MTTD, MTTR — correctly shows 0.0 and uses the unit field to show 'min' or 'hrs'.
MTTD: was querying AuditLog for action names that don't match actual
logged actions. Now uses red_started_at → blue_started_at directly
(both stored on the Test record). Net of red_paused_seconds.
MTTR: was searching for remediation_status=completed (no data). Redefined
as total pipeline time: red_started_at → blue_validated_at net of all
paused time. Only counts fully validated tests.
Red avg time: was using red_validated_at - created_at (created_at NULL
for many tests). Now uses blue_started_at - red_started_at net paused.
Blue avg time: was using blue_validated_at - red_validated_at (wrong
phase boundary). Now uses blue_work_started_at (or blue_started_at
fallback) → blue_validated_at net of blue_paused_seconds.
- Convert horizontal bar chart to vertical bars (columns)
- Sort all 14 MITRE ATT&CK tactics in official order:
Reconnaissance → Resource Development → Initial Access → Execution →
Persistence → Privilege Escalation → Defense Evasion → Credential Access →
Discovery → Lateral Movement → Collection → C2 → Exfiltration → Impact
- Show ALL tactics (not a subset)
- Labels rotated -45° to fit all names
- Bars have rounded top corners; horizontal gridlines only
'Validation Throughput (tests/week)' was time-dependent — director wanted
an activity-based metric instead.
New metric: Pipeline Conversion Rate
formula: validated / (validated + rejected + in_review) × 100
unit: % (no time reference)
meaning: 'of all tests that have entered validation, X% succeeded'
trend: declining if in_review backlog > validated count,
improving if conversion ≥ 80%, stable otherwise
Backend: calculate_validation_throughput() rewritten — same API key
(tests_per_week) kept for compatibility, new conversion_rate field added.
Frontend: label → 'Pipeline Conversion', unit → '%', tooltip updated.
Replace single list with two-column layout:
- LEFT '⚠ Highest Exposure': top 5 actors by uncovered technique count,
red border, text explaining 'these attacks would go unnoticed today'
- RIGHT '✅ Strongest Detection': top 5 actors by coverage %, green border,
text explaining 'Blue Team would likely detect an intrusion from these'
Shows both the risks (where to focus testing) and the strengths
(what's already well protected) to give executives a balanced view.
New MetricTooltip component — a small ⓘ icon showing an executive-
friendly explanation panel on hover (CSS, no JS, instant).
DashboardPage: tooltips on all 6 coverage summary cards (Total
Techniques, Validated, Partial, In Progress, Not Covered, Not
Evaluated), Coverage Evolution chart, Test Pipeline funnel,
Team Activity and Validation Rate section headers.
ExecutiveDashboardPage: tooltips on all 4 sub-scores (Coverage,
Detection, Critical, Response), Score Trend, Top Threat Actors,
4 KPIs (MTTD, MTTR, Detection Efficacy, Validation Throughput),
Coverage by Tactic, Critical Gaps table, and all 6 team metrics
(Red/Blue Tests Done, Avg Time, Rejection).
Each tooltip explains what the metric measures, what a good/bad
value looks like, and what action to take — written for non-
technical executives.
Tempo rejects durations under 60 seconds ('Duration must be at least
one minute'). Now:
- Always send ≥ 60 s (1 minute minimum)
- Round UP to nearest whole minute (math.ceil)
- 2 s → 60 s, 3m20s (200s) → 240 s, 5m00s (300s) → 300 s
red_tech can only edit procedure/tool/summary when the test is in
red_executing state (after pressing Start Execution). In draft state they
see a read-only view and an orange hint 'Press Start Execution to begin
editing — the timer must be running first.'
blue_tech can only edit when blue_work_started_at is set (after pressing
Start Evaluation). Before that they see an indigo hint 'Press Start
Evaluation to begin editing — pick up the test first.'
red_lead, blue_lead and admin are unaffected — they retain full edit
access in all applicable states including draft.
Problem: 15-minute tokens with no refresh mechanism kicked users to login
even when actively using the app.
Fixes:
1. config.py: raise ACCESS_TOKEN_EXPIRE_MINUTES from 15 → 480 (8h).
Reasonable for an enterprise internal tool; still configurable via env.
2. POST /auth/refresh: new endpoint that reads the current aegis_token
cookie and issues a fresh token if the session is still valid. Returns
the new token in the cookie + body (same shape as /auth/login).
3. frontend/api/client.ts: response interceptor now attempts a silent
refresh on 401 before redirecting to login:
- Calls POST /auth/refresh once per failed request
- If refresh succeeds: retries the original request transparently
- If refresh fails: redirects to /login as before
- Deduplicates concurrent refresh attempts (refresh once, resolve all)
- Never attempts refresh on /auth/refresh or /auth/login themselves
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Root cause: Microsoft Teams Incoming Webhooks require MessageCard JSON
format. The service was sending generic Aegis JSON which Teams rejected
with a 400, incrementing failure_count on every dispatch.
Fix: _send_webhook() now auto-detects the target from the URL:
- webhook.office.com / teams.microsoft.com → Teams MessageCard
(colored card with event title + key/value facts table)
- hooks.slack.com → Slack attachments format
- everything else → current generic Aegis JSON
Also resets failure_count=0 in production so the webhook starts fresh.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Backend:
- submit_red_evidence: raises InvalidOperationError if no Red Team
evidence file has been uploaded for the test
- submit_blue_evidence: raises InvalidOperationError if no Blue Team
evidence file has been uploaded
Frontend:
- 'Submit to Blue Team' button: disabled + '⚠ Upload evidence first'
hint when test.red_evidences is empty
- 'Submit for Review' button: same for test.blue_evidences
- Native tooltip on disabled buttons explains the requirement
- Buttons re-enable automatically after the first file is uploaded
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Backend: GET /campaigns/{id}/timing-summary
Aggregates timing across all campaign tests:
- red_execution_secs: red_started_at → blue_started_at (minus paused)
- blue_queue_secs: blue_started_at → blue_work_started_at
- blue_evaluation_secs: blue_work_started_at → validated (minus paused)
- total_secs: sum of all three phases
Returns totals + per-test breakdown sorted by total time desc.
Frontend: new CampaignTimingPanel component replaces WorklogTimeline
- 4 summary cards: Red Execution / Blue Queue / Blue Evaluation / Total
- Stacked horizontal bar showing time distribution
- Per-test breakdown with individual mini-bars and phase durations
- Shows 'No tests started yet' when no timing data available
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
title= attribute tooltip is browser-native, tiny, and often invisible.
New StatusBadge component uses a Tailwind group-hover absolute panel
that appears immediately on hover with:
- Clear heading per status
- 'Meaning' and 'Action' lines
- Arrow pointing to the badge
- 200ms fade-in transition
Used in TechniquesPage (list table) and TechniqueDetailPage (header).
1. Status logic (v3): require ≥2 validated tests with 'detected' result
to reach 'validated' status. With only 1 validated+detected test the
technique stays 'partial' (single test is insufficient evidence).
Backfilled existing data: T1012 and T1059.001 downgraded to 'partial'.
2. Hover tooltips on status badges in TechniquesPage and TechniqueDetailPage:
- validated: ≥2 tests executed and detected
- partial: some tests done but incomplete coverage
- in_progress: tests exist but none validated yet
- not_covered: tests run but Blue Team didn't detect
- not_evaluated: no tests created yet
- review_required: recent update needs acknowledgment
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Previously: alphabetical order (first 5 actors from list_actors query).
Now: ranked by uncovered technique count = technique_count × (1 - coverage_pct/100).
Tiebreak: higher technique_count first (broader attack surface).
Fetches 100 actors, sorts client-side, shows top 5 with:
- Rank badge (1-5) colored red/orange/yellow/gray
- 'N uncovered / M techniques' subtitle instead of target sectors
- Coverage bar + percentage
This ensures the actors with the largest coverage gap appear first.
Root cause: Cloudflare CDN was caching empty/error API responses from
/api/v1/metrics/* endpoints during the backend startup window (502 errors).
Subsequent requests were served from Cloudflare edge cache, never reaching
nginx or the backend, so the dashboard always showed empty metrics data.
NoCacheAPIMiddleware adds Cache-Control: no-store + Pragma: no-cache to
all /api/ responses so Cloudflare and browsers never cache them.
Root cause: after backend restart (502 errors on startup), metric queries
(pipeline, team, recent, validation) get cached in error state. When the
user stays on the dashboard, the component never remounts so queries don't
auto-retry.
Fixes:
1. refetchOnMount:'always' — queries ALWAYS refetch when component mounts,
even if cached with error/stale data. Prevents stuck empty state.
2. gcTime:0 — error state is not cached; next mount starts a fresh query.
3. retry:3 — more retries before giving up (covers slow startup windows).
4. Refresh button in header — manually invalidates and refetches all 4
metric queries with a single click. Spinner icon during refetch.
1. metrics_query_service: use NULLS LAST in get_recent_tests() so tests
with actual dates always appear before NULL-dated ones.
2. campaign_service: set created_at=datetime.utcnow() when creating tests
from campaigns (was missing, leaving 21 tests with NULL created_at).
Fixed existing NULL values directly in production DB.
3. DashboardPage: add isError handling to all V2 metric widgets
(pipeline, team activity, validation rate, recent tests).
- Add retry:2 to all secondary metric queries so transient failures
are retried before showing empty state.
- Show 'Could not load X — refresh' instead of empty/misleading
'No tests created yet' when a query actually fails.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
1. Duplicate function definition: the old 2-param _entry_matches shadowed
the new 3-param version — Python uses the last definition, so the call
with 3 args threw TypeError. Removed the stale old definition.
2. NIST NVD deprecated their XML RSS feeds in 2023 — URL returns 404.
Replaced with SecurityWeek RSS which is active and covers CVEs/threats.
RT tests are created in 'in_review' state (not validated):
- red_validation_status = 'approved' (RT confirmed execution)
- blue_validation_status = null (pending Blue Lead review)
- detection_result is pre-filled from the import JSON
Blue Lead sees these in their normal validation queue and confirms
or rejects the detection result. Only after Blue Lead approval does
the technique coverage update to validated/not_covered/partial.
This gives Blue Lead oversight over RT findings rather than auto-
accepting external engagement results as ground truth.
Backend — POST /tests/import-rt (red_lead + admin):
Accepts engagement JSON with name/date/description/operator and
a list of techniques each with mitre_id, result, attack_success,
platform, notes. Creates one Test per technique directly in
'validated' state (red + blue validation = approved) bypassing
the normal workflow. Recalculates technique.status_global for
all affected techniques. Returns created/skipped summary.
Frontend — /tests/import-rt (new dedicated page):
- Format reference panel (collapsible) with field descriptions
- Download template JSON button (generates a filled example)
- Paste JSON textarea + file upload (.json)
- Live validation + preview table showing what will be imported
- Import button with spinner
- Success / warning / error result display
Accessible to admin and red_lead only.
Added to sidebar under Tests > Import RT Results.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Backend:
- intel_service: remove 50-technique limit (scan all techniques), improve
pattern matching with word boundaries (\bT1059\b), raise min name length
to 8 chars to reduce false positives, skip entries with empty titles
- technique_query_service: add intel_items to get_technique_detail() so
the technique page now shows recent threat intel articles (last 20)
- New GET /intel/items endpoint with optional technique_id filter
Frontend:
- New api/intel.ts with listIntelItems()
- ReviewQueuePage: complete redesign
* Expandable rows — click a technique to see its intel articles inline
* IntelPanel component fetches articles per technique on expand
* 'Create Template from Intel' button opens pre-filled modal:
name (from article title), source_url (article link), technique_id
User reads the article and fills the attack procedure
* Updated explanation text: lists all 3 reasons a technique can be flagged
(MITRE update / intel scan / new template or detection rule)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
TestCatalogPage: 'Use Template' button had no role check — any user
(including viewer/blue_tech/red_tech) could see and click it, which
would fail at the backend (POST /tests/from-template requires
red_lead|blue_lead). Added canUseTemplate check; button hidden for
viewer, blue_tech, red_tech.
TechniqueDetailPage: 'Run This Test' / 'Re-run' buttons in the
Available Templates section also had no role check. Added canRunTemplate
(same criteria: admin|red_lead|blue_lead). The 'View test' button for
active tests remains visible to everyone (read-only navigation).
Principle: if a user cannot perform the action, the button does not
appear — no permission error messages, just absence of the control.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
1. /executive-dashboard: add 'viewer' to ProtectedRoute roles — sidebar
showed the link to viewers but the route redirected them to /dashboard.
2. /comparison: same fix — viewer was in sidebar roles but not in route.
3. /techniques/review-queue: add ProtectedRoute (leads+admin) — the page
had no route-level protection, any authenticated user could access it.
4. TechniqueDetailPage review banner: hide from users who can't act on it.
Previously shown to everyone with a 'Leads only' badge; now only shown
to canReview users (admin/red_lead/blue_lead). Non-leads don't need to
see alerts about changes they cannot acknowledge.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
MITRE ATT&CK STIX data never includes primary_motivation on intrusion-set
objects. Motivation is now derived with a 3-tier fallback:
1. Curated MITRE-ID override map (100+ known groups mapped by hand)
2. STIX primary_motivation field (if MITRE ever adds it)
3. Description keyword inference (financ/ransomware/espionage/
nation-state/destructive/hacktivist patterns)
Re-running MITRE sync will now backfill motivation for existing actors.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
1. fix(search 500): func.cast(col, func.text()) is invalid SQLAlchemy —
replaced with cast(col, Text) for both aliases and target_sectors
JSONB columns. Generating correct CAST(col AS TEXT) SQL.
2. feat(motivation): extract primary_motivation and sophistication from
STIX intrusion-set objects during MITRE sync. Added _normalize_motivation()
to map STIX vocabulary → simplified frontend values (espionage / financial /
destruction / hacktivism). Both create and update paths now set these fields.
Run MITRE sync to backfill existing actors.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Implements the Digital Operational Resilience Act as a compliance framework
using the same pattern as CIS Controls v8 (hardcoded curated mappings,
no official STIX bundle exists for DORA).
22 controls across 5 chapters mapped to MITRE ATT&CK techniques:
Ch. II — ICT Risk Management (Art. 5–15): governance, identification,
protection, detection, response, backup, threat intel
Ch. III — Incident Management (Art. 17–19): classification, reporting
Ch. IV — Resilience Testing (Art. 24–27): general testing + TLPT
(Art. 26 explicitly based on TIBER-EU/ATT&CK threat-led testing)
Ch. V — Third-Party Risk (Art. 28, 30, 42): supply chain, trusted rels.
Ch. VI — Information Sharing (Art. 45)
Technique mappings derived from ENISA DORA guidelines and TIBER-EU framework.
Import is triggered via POST /api/v1/compliance/import/dora (admin only).
Frontend: new 'DORA' button in the Compliance page import section.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The useState hook was placed after the isLoading/error early returns,
violating the Rules of Hooks. First render hit the early return without
calling the hook; second render (after data loaded) called it, producing
'more hooks than previous render' — React error #310 and a white screen.
Moved const [showTemplateModal] to the state block at the top of the
component, alongside the other useState declarations.
Previously a JS rendering error produced a blank white screen with no
feedback. PageErrorBoundary now catches the error, shows the error
message + stack trace, and offers a reload button. This will surface
the exact crash message for the inaccessible test page.
Adds a 'Save as Template' button in the Details sidebar (visible to
red_lead, blue_lead and admin only). Opens a modal pre-filled from
the test's own fields:
test.name → template name
test.description → description
test.platform → platform
test.procedure_text → attack_procedure
test.tool_used → tool_suggested
test.technique_mitre_id → mitre_technique_id
User can also set severity and write expected_detection (Blue Team
guidance — not stored on tests). Calls POST /test-templates with
source='custom' on submit.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
/techniques (technique browser with filters) was an orphaned route —
only reachable via 'Back to techniques' button or direct URL.
Now exposed in the sidebar as part of a new ATT&CK group:
ATT&CK ▾
Techniques → /techniques
Coverage Matrix → /matrix
Review Queue → /techniques/review-queue (leads+admin only)
Child role filtering added to SidebarLink.
Review Queue badge moved to the ATT&CK group header.
CSS grid was stretching both cells in a row to the same height as the
tallest card, making the unexpanded card appear 'open' with blank space.
items-start makes each card only as tall as its own content.
def.id was undefined for D3FEND items, causing expandedId===undefined
to match ALL items simultaneously and opening every card at once.
d3fend_id (e.g. 'D3-DSDP') is always present and unique per defense.
Techniques don't have their own Jira tickets — tickets exist on tests
and campaigns. The previous JiraLinkPanel entityType='technique' always
returned empty.
Backend: add entity_ids (list) filter to GET /jira/links so multiple
test IDs can be fetched in a single request.
Frontend API: listJiraLinks() accepts entity_ids[] and serialises them
as repeated query params (required by FastAPI List[UUID] parsing).
TechniqueDetailPage: replace JiraLinkPanel with TechniqueJiraSection —
a dedicated read-only component that:
- Takes technique.tests (already loaded)
- Batch-fetches all test Jira links in one request
- Shows test name + ticket key + status + priority + open-in-Jira link
- Hides itself when no tickets exist (avoids empty panel)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
WorklogTimeline: add readOnly prop — hides 'Log Time' button and form.
TestPhaseTimeline: remove 'Sync to Tempo' button from TempoSyncBadge;
only displays the green 'Tempo' badge when already synced. Cleans up
unused imports (useState, useMutation, useQueryClient, syncTestToTempo).
CampaignDetailPage: JiraLinkPanel and WorklogTimeline both now rendered
with readOnly=true; JiraLinkPanel receives campaign name as label.
Jira tickets and time worklogs are created automatically by the system
(campaign activation, test workflow) — no manual editing from detail pages.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Extends the review queue triggers to cover test template imports:
- atomic_import_service: flags techniques when new Atomic Red Team
templates are imported
- caldera_import_service: same for Caldera templates
- lolbas_import_service: same for LOLBAS templates
- test_templates router (manual creation): flags the technique when
an admin/lead creates a custom template via the API
Pattern is identical to the Sigma/Elastic detection rule approach:
collect new mitre_ids during the loop, bulk-update after commit.
Manual creation does a single technique lookup and sets the flag
inside the existing UnitOfWork.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
1. Campaign test deletion: removing a test from a campaign now also
deletes the underlying Test record and recalculates technique status.
2. Review Queue triggers: review_required=True is now also set when
- Sigma/Elastic detection rules are imported for a technique
- A test is validated (coverage status changes)
3. Test detail — Technique link: 'Technique' entry added at the top of
the Details sidebar showing MITRE ID + name as a clickable link to
/techniques/{mitre_id}.
4. Jira panel — read-only on test page: added readOnly + label props to
JiraLinkPanel. TestDetailPage now passes readOnly=true and the test
name as label, hiding Link Issue / Sync / Unlink controls (automatic
Jira creation only — no manual management).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Each template card in 'Available Test Templates' now shows contextual
status derived from technique.tests (already loaded):
- Active test (draft/executing/evaluating/in_review):
blue 'Executing / In Review' badge + 'View test →' button
(prevents blind duplicate creation)
- Validated / detected (fresh):
green 'Detected' badge + dimmed 'Re-run' button
- Validated / not_detected or partial:
red/yellow result badge + full 'Run This Test' button (re-run encouraged)
- Validated but stale (review_required=true):
result badge + '⚠ Coverage may be stale' line
- No tests: normal 'Run This Test' button
No extra API calls — status is derived from the technique detail
already in-memory.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Backend: add not_in_any_campaign filter to list_tests (subquery on
CampaignTest) and expose it as a query param on GET /tests.
Frontend: the 'Existing Test' tab now requests only
state=draft & not_in_any_campaign=true
so tests already linked to any campaign or not in draft state
are never shown.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- From template: name is pre-filled as '[Campaign] {template.name}'
(user can edit before confirming).
- Existing test: renamed via PATCH /tests/{id} to prepend '[Campaign] '
before being linked to the campaign, consistent with the APT-generated
campaign flow.
Idempotent — skips rename if the name already starts with '[Campaign]'.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The modal now has two tabs:
- 'From Template' (default): searchable/filterable template catalog
→ select template → customise name/platform/procedure/tool
→ 'Create & Add to Campaign' (two-step: POST /tests/from-template
then POST /campaigns/{id}/tests)
- 'Existing Test': previous behaviour — add an already-created test
Both tabs share an added-count footer badge.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- New /techniques/review-queue page: lists all techniques flagged for
review after a MITRE ATT&CK sync, grouped by tactic. Leads and admins
can mark each one reviewed inline without leaving the page.
- Sidebar: 'Review Queue' link (admin/red_lead/blue_lead only) with an
amber badge showing the live pending count.
- TechniqueDetailPage: amber banner when review_required=true explaining
what happened and who can act; 'Mark as Reviewed' button now amber
coloured for visual distinction. 'Leads only' chip shown for blue_tech.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The tick/cross buttons navigated to /tests/:id/validate and /tests/:id/reject
which are non-existent routes (catch-all redirected to dashboard).
Removed both buttons; the View (FileText) icon is the correct entry point
to the test detail page where the full workflow lives.
(Citation: ...) patterns are stripped from body text, replaced with
Unicode superscript numbers (¹²³), and shown in a compact "Sources"
section below — collapsed when there are more than 3, expanded otherwise.
Deduplication ensures the same citation reference appears only once.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- New shared MarkdownText component (react-markdown + remark-gfm)
that renders links, bold, italic, lists, code, blockquotes.
External links open in a new tab with rel=noopener.
- Applied to: technique description, threat actor description,
test description, campaign description, detection rule descriptions,
D3FEND defense descriptions, red/blue summaries and validation notes.
- procedure_text (code/commands) stays in <pre> — not processed as MD.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Sidebar: add `end` prop to child NavLinks so "All Tests" (/tests) is
only highlighted when exactly on /tests, not on /tests/validated.
- Backend: recalculate technique status_global for all affected techniques
when tests are deleted via delete_campaign(delete_tests=True), preventing
stale coverage metrics on the dashboard.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- New /tests/validated page with its own route and sidebar link, showing
only validated tests with Attack and Detection result badges.
- Removed the duplicate "My Pending Tasks" sidebar entry (same as All Tests).
- All Tests table no longer shows validated tests; clicking the Validated
counter card navigates to the new page instead.
- Validated option removed from the state filter dropdown in All Tests.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Validated tests no longer appear in the active tests table; they are
shown in a dedicated collapsible "Validated Tests" section at the bottom
(with its own sortable table and count badge).
- Added "Waiting" column to the main table showing elapsed time since
last update for blue_evaluating tests, sortable so Blue Team can
prioritise the oldest pending evaluations.
- Sorting by Waiting pushes blue_evaluating rows to the top and orders
them oldest-first by default.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
When must_change_password is true the user must pick a genuinely new
password. Added a verify_password check against the existing hash before
accepting the new value, raising BusinessRuleViolation if they match.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The form captured name/description/platform/procedure/tool edits but
never sent them — the created test always used the raw template values.
- TestTemplateInstantiate schema: add optional override fields
(name, description, platform, procedure_text, tool_used)
- create_test_from_template service: accept *_override kwargs;
use override value when provided, fall back to template value
- Router: pass all override fields from payload to service
- Frontend API createTestFromTemplate: accept overrides object, spread into body
- TestFromTemplateForm: pass all form state values as overrides
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Backend:
- technique_query_service.get_technique_detail() now queries DetectionRule
by mitre_technique_id == mitre_id (same field the heatmap uses)
- Rules sorted: critical → high → medium → low → informational, then alphabetically
- Returns: id, title, description, source, source_url, rule_format,
severity, platforms, false_positive_rate
Frontend:
- New DetectionRulesSection component with expandable rows per rule
- Color-coded severity dots and badges (red/orange/yellow/blue/gray)
- Source badges (sigma=purple, elastic=blue, splunk=orange, custom=cyan)
- Shows format, false positive rate, platforms, source link on expand
- Empty state when no rules exist
Fixes: T1189 showed green in heatmap but no rules on detail page
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Auto-trigger POST /risk/compute on first load if no profiles exist
- Add "Refresh scores" button next to Critical Gaps header (spins while computing)
- Add computeRiskScores() to frontend/src/api/risk.ts
- After compute, invalidate risk-profiles query so table updates immediately
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Create frontend/src/api/risk.ts with getRiskProfiles() API function
- Executive Dashboard fetches risk profiles and builds a techniqueId→profile map
- Critical Gaps sorted by risk_score DESC (highest risk shown first)
- Ties resolved: not_covered before not_evaluated; unscored techniques last
- Table now shows Risk Score (0-100, color-coded) and Risk Level badge per row
- Column renamed to "Critical Gaps — Top 10 by Risk Priority"
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Backend: add DELETE /campaigns/{id}?delete_tests=bool endpoint
- Backend: add delete_campaign() service — handles draft-only restriction,
optional test deletion, nullifies child campaign FKs
- Backend: remove early Jira ticket creation from POST /campaigns,
POST /campaigns/{id}/tests, and POST /campaigns/from-threat-actor
- Backend: activate endpoint now creates campaign Jira ticket if missing,
then creates test tickets (all deferred from creation to activation)
- Frontend: add deleteCampaign() API function to campaigns.ts
- Frontend: two-step confirmation dialog on CampaignDetailPage —
first confirms deletion, then asks whether to also delete associated tests
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Remove WorklogTimeline (manual time log) from test detail page
- TestPhaseTimeline now accepts testId, fetches its own worklogs,
and shows Tempo sync status on the Red Team Execution row:
• green badge if already synced (with worklog ID tooltip)
• 'Sync to Tempo' button (blue) if not yet synced
- Add POST /tests/{id}/sync-tempo backend endpoint for manual sync:
finds unsynced red_team_execution worklogs and pushes them to Tempo
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Adds a View button (eye icon) on each evidence card for previewable file
types. Opens a full-screen modal:
- Images (png/jpg/gif/webp/svg/…): rendered directly via <img> tag
- JSON: fetched authenticated, pretty-printed in green mono
- Text/log/md/csv/xml/yaml/…: fetched authenticated, shown in <pre>
Non-previewable files only show the Download button as before.
Modal closes on Escape or backdrop click.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
FastAPI 0.136.1 + Pydantic 2.13.4 serialises responses via TypeAdapter which
calls the compiled Rust validator directly, bypassing any Python-level
`model_validate` classmethod override. The @model_validator(mode='before')
decorator IS invoked by the Rust pipeline, so the evidence red/blue split and
technique field population now run on every serialisation path.
Also eager-load technique in get_test_detail to avoid lazy-load surprises.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Root causes found for Tempo worklogs never reaching Tempo:
1. Wrong API region: workspace is on api.eu.tempo.io/4 but code used api.tempo.io/4
→ Tempo returned "User is invalid" (400) for all POST /worklogs
2. Trailing space in jira_account_id stored in DB (now stripped with .strip())
3. tempo_synced field was never updated even on success (now set from Tempo response)
Fix: add tempo.base_url system_config key (admin-configurable without redeploy),
fall back to TEMPO_BASE_URL env-var, then global default. DB already updated with
https://api.eu.tempo.io/4 for this workspace.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Accessing obj.evidences on a session-expired ORM object (mutation endpoints
do commit+refresh without joinload) triggers a lazy query that fails or
returns stale data. Use obj.__dict__.get('evidences') instead — SQLAlchemy
stores joinloaded relationships in __dict__; absent means not loaded.
Mutation endpoints (submit-red, submit-blue, etc.) return empty evidence
lists, which is fine: the frontend invalidates and refetches GET /tests/{id},
which uses joinedload and correctly populates red_evidences / blue_evidences.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
tempo: tempoapiclient raises SystemExit (BaseException) on API errors like
'User is invalid' 400 responses; except Exception never catches it, killing
the uvicorn worker and causing a 500. Wrap create_worklog() to intercept
BaseException and re-raise as RuntimeError so callers can catch it safely.
evidence: TestOut schema was missing red_evidences / blue_evidences fields.
The ORM model has evidences loaded via joinedload but they were never
serialized into the API response. Add both fields to TestOut and override
model_validate to split Test.evidences by team, injecting the backend-proxy
download_url for each one (/api/v1/evidence/{id}/file).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- tempo: remove unsupported `workType` kwarg from create_worklog call;
tempoapiclient v4 does not accept it → was causing every Tempo sync to fail
- tests: set created_at=datetime.utcnow() explicitly on test creation (both
create_test and create_test_from_template) since the DB column has no
server default, causing 'Created —' in the UI
- jira: remove duplicate Proof of Concept section from ticket description body;
PoC already lives in customfield_10309, no need to repeat it in description
- ui: add TestPhaseTimeline component (read-only) showing RT execution time,
blue queue time, blue evaluation time and lead validation timestamps derived
from test phase timestamps; placed above WorklogTimeline in test detail page
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Evidence download:
- Replace presigned MinIO URLs with backend proxy endpoint
GET /api/v1/evidence/{id}/file streams the file through the backend
so MinIO never needs to be publicly accessible from browsers
- Add download_file() helper to storage.py (internal boto3 get_object)
- download_url in EvidenceOut now points to the proxy endpoint
Jira attachment:
- Fix add_attachment call: use add_attachment_object(issue_key, BytesIO)
instead of add_attachment(issue_key, filename=..., content=...) which
had wrong keyword args for the installed atlassian-python-api version
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Jira — PoC custom field:
- Add customfield_10309 (Proof of Concept) to issue fields when creating
test tickets so the attack procedure appears in the dedicated Jira field
Tempo — blue team exclusion:
- Remove blue_team_evaluation from _TEMPO_ACTIVITY_TYPES; blue team time
is tracked internally (worklogs) for SLA but never sent to Tempo since
blue team has no Jira access
Evidence — uploaded_at NULL fix:
- Set uploaded_at=datetime.utcnow() explicitly in upload_evidence router;
the DB column has no server default so it was saving as NULL
Evidence — presigned URL browser access:
- Add MINIO_PUBLIC_ENDPOINT setting (config.py, docker-compose.prod.yml)
- storage.py uses a dedicated _public_client for presigned URL generation
so browsers receive URLs with the publicly accessible hostname instead of
the internal Docker service name (minio:9000)
- Expose MinIO port 9000 in docker-compose.prod.yml
Evidence — Jira attachment:
- After upload to MinIO, call jira.add_attachment() to attach the file to
the linked Jira ticket (non-fatal; errors are logged and swallowed)
Settings — hide Jira/Tempo from blue team:
- ProfileSection checks user role; blue_lead and blue_tech do not see the
Jira Integration or Tempo Integration personal settings sections
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Campaign issue type changed from Task to Epic (required to nest under
Initiative OFS-20795 in classic Jira)
- Added customfield_10011 (Epic Name) — required when creating Epics
- Removed JIRA_ISSUE_TYPE_SUBTASK; all tests are now Task regardless of
whether they are inside a campaign or standalone
- Standalone tests use the configured standalone parent (OFS-20798, an
Epic) so Task→Task parent is never attempted
- Campaign tests use the campaign Epic key passed via parent_ticket_override
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
OFS-20798 is a Task (child of OFS-20795 Epic), so tests nested
under it must be Sub-tasks, not Tasks — Task cannot parent Task.
Logic:
- parent_ticket_override (campaign) → Sub-task (unchanged)
- standalone_parent configured and differs from general parent → Sub-task
- only general parent (Epic) → Task
This fixes 'Please select valid parent issue' for standalone tests.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The field is the general parent (e.g. OFS-20795) under which campaigns
are created directly. 'Campaign Parent Ticket' was misleading.
Standalone Tests Parent Ticket remains separate (e.g. OFS-20798).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
1. Jira status → In Progress on Start Execution
- push_test_event calls set_issue_status("In Progress") when
new_state == "red_executing" (non-fatal, separate try/except)
2. Jira assignee set on Start Execution
- assign_issue() called with actor.jira_account_id when operator
clicks Start (non-fatal)
3. Standalone tests parent ticket (OFS-20798)
- New jira.parent_ticket_standalone config key
- get_jira_parent_ticket_standalone() falls back to parent_ticket
- auto_create_test_issue uses standalone parent for non-campaign tests
- Exposed in /system/jira-config GET+PATCH and SettingsPage UI
4. Tests table: Created + Updated columns
- Add Created column (created_at), fix Updated to show updated_at
- Both use UTC-aware date parsing (append Z if no tz suffix)
- updated_at added to Test TypeScript interface
5. Sortable columns in tests table
- All 7 columns sortable: Name, Technique, State, Current Team,
Platform, Created, Updated
- Click to sort asc, click again to reverse; ChevronUp/Down indicator
- Default sort: Created desc (newest first)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Backend returns naive UTC datetimes without 'Z' suffix. JavaScript
new Date("2026-05-27T09:29:18") parses as local time (UTC+2 in Spain),
making the timer start at 02:00:06 instead of 00:00:00.
Fix: append 'Z' to any timestamp string that lacks timezone info before
passing it to new Date(), so the browser always interprets it as UTC.
Applied to both startedAt and pausedAt in LiveTimer.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Previously blue_started_at was set when the RED team submitted evidence
(= queue open time), so Tempo was getting total queue wait time instead
of actual work time.
Changes:
- DB: add blue_work_started_at column (migration b045), set when a blue
tech explicitly picks up the test (mirrors red_started_at for red team)
- Workflow: new start_blue_work() function + POST /tests/{id}/start-blue-work
endpoint (blue_tech / blue_lead roles). Cannot be called twice.
- submit_blue_evidence: uses blue_work_started_at (when available) as the
phase start for the Tempo worklog, falls back to blue_started_at
- reopen_test: clears blue_work_started_at alongside other timing fields
- Tempo: both red_team_execution and blue_team_evaluation now synced;
correct work_date and description per activity type
- Frontend: "Start Evaluation" button shown in blue_evaluating state when
blue_work_started_at is null; live timer shows from pick-up time
What each timestamp tracks:
blue_started_at = queue entry (SLA / internal tracking)
blue_work_started_at = pick-up by blue tech (Tempo start)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Two bugs fixed:
1. Blue team evaluation was also sent to Tempo. Only operator (red team)
execution time should be logged — blue team time is tracked internally
in Aegis but does NOT represent billable operator work. Added a
whitelist (_TEMPO_ACTIVITY_TYPES = {"red_team_execution"}).
2. _calculate_duration() re-computed duration from red_started_at to
datetime.utcnow() at call time, without subtracting paused seconds.
This caused inflated times (e.g. 45 min instead of 5 min) when there
was any delay between the workflow transition and the Tempo call.
Now the duration_seconds already computed by _create_phase_worklog
(gross elapsed - paused) is passed directly to auto_log_test_worklog
and used as-is, so Aegis and Tempo always agree on the duration.
Also: use red_started_at as the worklog date (not submission timestamp)
so the Tempo entry reflects when the work actually happened.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
get_worklogs_by_account_id does not exist in tempoapiclient v4.
The correct method is search_worklogs(dateFrom, dateTo, authorIds=[...]).
Also improve error messages: 401 points to where to get the token,
404 tells the user the Account ID may be wrong.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Root cause: Jira rejects Task-under-Task nesting ("Please select valid
parent issue"). Campaign tickets and test tickets were both created as
Task, so nesting test under campaign failed for all 62 APT32 tests.
Fix:
- JIRA_ISSUE_TYPE_CAMPAIGN: "Epic" -> "Task" (was unused, now used)
- JIRA_ISSUE_TYPE_SUBTASK: "Sub-task" (new config key)
- auto_create_campaign_issue: uses JIRA_ISSUE_TYPE_CAMPAIGN (Task)
- auto_create_test_issue: uses Sub-task when parent_ticket_override is
set (campaign context), Task otherwise (standalone)
Hierarchy: OFS-9107 -> Campaign (Task) -> Test (Sub-task)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
When a campaign is activated (Start), iterate all its tests and create
Jira tickets nested under the campaign ticket for any test that doesn't
already have one. Mirrors the pattern used in generate_campaign_from_actor.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Each user can now store their own personal Tempo API token in their
profile settings. Time is logged using each user's own credentials.
Backend:
- Migration b044: adds tempo_api_token column to users table
- User model: adds tempo_api_token column
- UserPreferencesUpdate: adds tempo_api_token field (write-only)
- UserOut: adds tempo_api_token (excluded) + tempo_token_set bool;
@model_validator derives both jira_token_set and tempo_token_set
- users router: handles tempo_api_token same as jira_api_token
(empty string clears it, never returned in responses)
- tempo_service: refactored to per-user token; has_tempo_configured(),
get_user_tempo_client(user) use user.tempo_api_token; global
TEMPO_ENABLED still acts as kill-switch
- system router: /system/tempo-test now uses current user's personal
token (any role); removed global TEMPO_API_TOKEN dependency
Frontend:
- settings.ts: UserPreferencesUpdate.tempo_api_token, UserMeOut.tempo_token_set
- SettingsPage ProfileSection: Tempo Integration section with password
field, show/hide toggle, configured badge, and Test Tempo button —
mirrors the Jira token UX exactly
- JiraConfigSection: removed stale global Tempo test block
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Jira URL fix:
- JiraLinkPanel now fetches the configured Jira base URL via getJiraConfig()
instead of hardcoding https://jira.atlassian.com; falls back to the old
value if config is not yet loaded
Description fix:
- _build_test_description: renamed 'h3. Procedure' -> 'h3. Proof of Concept'
so the procedure/tool block maps to the correct Jira field label
Tempo debug:
- New POST /system/tempo-test endpoint: checks TEMPO_ENABLED, token,
user jira_account_id, and makes a real API call; always returns HTTP 200
with status field (Cloudflare-safe)
- docker-compose.prod.yml: added TEMPO_ENABLED, TEMPO_API_TOKEN,
TEMPO_DEFAULT_WORK_TYPE env vars (default off, ready to enable)
- SettingsPage: added 'Test Tempo Connection' button in Jira admin tab
with clear feedback showing what's missing
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The OFS Jira project does not have the default Jira priority scheme
(Highest/High/Medium/Low/Lowest), causing a 'priority selected is invalid'
error on every ticket creation. Removing the priority field lets Jira use
the project default.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Jira tickets now follow the correct hierarchy:
OFS-9107 (system parent)
├── Standalone test ticket (unchanged — was already working)
└── Campaign ticket (NEW — created on campaign creation)
├── Test 1 ticket (NEW — created per test)
└── Test 2 ticket (NEW — created per test)
Changes:
- jira_service: add auto_create_campaign_issue() — creates campaign
ticket as child of OFS-9107; stores JiraLink(entity_type=campaign)
- jira_service: add get_campaign_jira_key() / get_test_jira_key()
helpers to look up existing Jira links by entity
- jira_service: auto_create_test_issue() gains parent_ticket_override
param — when set, uses it as parent instead of OFS-9107
- campaigns router/create_campaign: triggers auto_create_campaign_issue
after commit
- campaigns router/from-threat-actor: triggers campaign ticket then
iterates campaign_tests and creates each test ticket under it
- campaigns router/add_test_to_campaign: if campaign has a Jira ticket
and the test has none yet, creates test ticket under campaign ticket
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- jira-test: when myself() returns empty displayName/emailAddress/name,
fall back to the configured Atlassian auth email so 'Connected as:' is
never empty
- jira-test: 401 error message now includes which email was used, making
misconfigured Jira email easier to diagnose
- jira-test: missing jira_url now returns HTTP 200 {status: error} instead
of HTTP 400, consistent with Cloudflare-safe pattern
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- jira-test now returns {status: "ok"|"error", message: ...} with
HTTP 200 so Cloudflare never intercepts the response
- jira_service strips trailing slash from URL before creating Jira
client (avoids double-slash in REST paths)
- Frontend reads data.status field instead of HTTP status code
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
FastAPI uses __pydantic_validator__.validate_python() which bypasses
model_validate() overrides. Switch to @model_validator(mode='after')
which the Pydantic Rust core always calls, so jira_token_set is now
correctly derived from the excluded jira_api_token field.
Also add a 10s timeout to the jira-test endpoint and better error
messages (the Atlassian library's "Expecting value" JSON error was
ambiguous).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Using setQueryData with the PATCH response means jira_token_set is
reflected in the UI instantly — no extra GET round-trip that could
leave the badge stale.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Replace render-body setState with useEffect so field initialisation
is idiomatic React and never races with user input. Also clarifies
placeholder text: empty token field = keep current, not clear it.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Users can now set a separate Atlassian email for Jira authentication
in Settings → Profile → Jira Integration. Falls back to the Aegis
account email when not set, so existing setups are unaffected.
- Migration b043: adds jira_email column to users table
- User model/schema: expose jira_email read/write
- jira_service: _effective_jira_email() uses jira_email ?? email
- Frontend: replaces read-only email display with editable input
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- backend: add parent_ticket field to JiraConfigOut/JiraConfigUpdate/_JIRA_KEYS
- backend: add get_jira_parent_ticket() helper in jira_service; use it in auto_create_test_issue() to set issue parent
- frontend/api: add jira_token_set to UserMeOut, jira_api_token to UserPreferencesUpdate, and full JiraConfigOut/Update types with getJiraConfig/updateJiraConfig/testJiraConnection functions
- frontend: expand ProfileSection with Jira API token password field (show/hide), token status badge, and account-id field
- frontend: add JiraConfigSection component (admin): enabled toggle, URL, project key, parent ticket, save + test connection
- frontend: add Jira tab (admin-only) with Link2 icon in SettingsPage sidebar
- Add jira_api_token field to User model + migration b042
- Per-user Jira client: user's corporate email + personal Atlassian token
- Admin-configurable Jira URL/project via system_configs (GET/PATCH /system/jira-config + POST /system/jira-test)
- Auto-create Jira ticket when a test is created (non-fatal)
- Push lifecycle comments on every state transition: draft→red_executing→blue_evaluating→in_review→validated/rejected→draft
- Rich ticket descriptions with technique, MITRE ID, priority from severity, labels
- UserOut.jira_token_set (bool) instead of exposing raw token
- PATCH /users/me/preferences now accepts jira_api_token
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Creates 14 comprehensive wiki pages covering architecture, roles,
test lifecycle, API reference, security, deployment, and QA guide.
Run from a machine with access to internal Gitea (192.168.1.107:3000).
- Accept 409 for playbook creation (unique per technique+type is correct behavior)
- Space logins 13s apart to avoid 5/min rate limit on login endpoint
- Reuse admin session from initial login to avoid duplicate login call
fix(qa): pass technique_id and test_id context between test suites
fix(qa): playbook creation requires technique_id field
fix(qa): lesson creation requires what_happened and root_cause fields
fix(qa): campaign complete test now activates with test before completing
fix(qa): rate limit test notes loopback exemption instead of failing
- fix(auth): enforce API key scopes in require_role/require_any_role;
attach _api_key_scopes to user on API key auth; add require_scope()
dependency — scopes were stored but never enforced (CWE-285)
- fix(sso): read SECURE_COOKIES env var for SSO cookie instead of
hardcoded secure=False — SAML sessions now respect HTTPS config (CWE-614)
- fix(webhooks): SSRF prevention — validate webhook URLs against private
and reserved CIDRs at creation/update time (CWE-918)
- fix(knowledge): restrict playbook/lesson create, update and restore
to admin/red_lead/blue_lead roles — was open to any authenticated user (CWE-284)
- fix(alerts): restrict alert acknowledge/resolve/dismiss to admin/lead
roles — any user could silence security alerts (CWE-284)
- security: delete get_admin_creds.py, check_auth.py, deploy.py scripts
containing hardcoded root SSH credentials and production DB access;
add scripts/.gitignore to prevent reintroduction (CWE-798)
- Add dispatch_webhook_targeted() to webhook_service for rule-specific delivery
- evaluate_all_rules() now dispatches in-app notifications (admins/leads) and
webhooks after each alert fires (targeted + global alert.fired broadcast)
- APScheduler: _run_alert_evaluation() job registered hourly alongside existing jobs
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
PostureSnapshot model, Alembic migration (b039exec), schemas, service
aggregating all phases (coverage/risk/operations/knowledge/MTTD), and
router at /api/v1/dashboard with executive view, KPIs, coverage-by-tactic,
posture-history, posture-snapshot, and activity-feed endpoints.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Playbooks: versioned Markdown runbooks per technique × type (attack/detect/investigate/respond/hunt)
- PlaybookVersion: immutable snapshots on every update; restore to any previous version
- LessonLearned: post-mortem records linked to tests/campaigns/attack-paths or manual
- Alembic migration b037know (raw SQL, idempotent, no PostgreSQL enums)
- Router /api/v1/knowledge: 14 endpoints for playbooks + lessons + stats
- Pydantic validators for playbook_type, severity, entity_type (422 on invalid)
- Knowledge stats endpoint: totals + breakdown by severity and playbook type
- Soft-delete on both resources; include_inactive filter for admin recovery
- QA script: 70+ tests across CRUD, versioning, filtering, auth, soft-delete, regression
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
POST /ownership/queue with an invalid reason or priority was silently
passing Pydantic and crashing at the DB layer (PostgreSQL enum type
mismatch → 500). Added @field_validator for both fields, matching the
existing validators in QueueItemPatch.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
SQLAlchemy fires before_create for ALL known enum types when any table
is created via op.create_table, causing DuplicateObject even with
create_type=False. Rewrite both CREATE TABLE statements as raw SQL via
conn.execute(sa.text(...)) and use CREATE TABLE IF NOT EXISTS / CREATE
INDEX IF NOT EXISTS for full idempotency.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Replace _enum_exists() helper (which had connection context issues in
Alembic) with PostgreSQL DO $$ BEGIN ... EXCEPTION WHEN duplicate_object
THEN NULL; END $$ blocks, which are truly idempotent regardless of
transaction state.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The previous fix changed _now() to return naive UTC, but the code still
called .replace(tzinfo=utc) on most_recent (from DB) before subtracting.
This caused "can't subtract offset-naive and offset-aware datetimes".
Now we strip tzinfo if present, keeping everything naive UTC consistently.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Replace datetime.now(timezone.utc) with datetime.utcnow() in _now() across
all three Phase 8 files to match DB DateTime column type (naive UTC)
- Guard POST /assets/{id}/techniques/{tid} against duplicate mappings:
if mapping already exists, update coverage_type/confidence_level instead
of inserting a duplicate row
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Phase 6.1: WebhookConfig model, CRUD router (/api/v1/webhooks, admin-only),
dispatch_webhook() with HMAC signing; integrated into test validation,
campaign completion, and MITRE sync job
- Phase 7.1: SMTP email service with send_test_validated_email,
send_campaign_completed_email, send_new_mitre_techniques_email;
notify_role_with_email() added to notification_service
- Phase 7.2: notification_preferences and jira_account_id on User model;
PATCH /users/me/preferences endpoint; Alembic migrations b031phase6 and b032phase7
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- nginx.conf: add new CSP script-src hash (sha256-Yvj83pg...) alongside previous one
- SystemPage: remove pencil icon from template name button, keep cyan underline style
- SystemPage: switch from selectedTemplate state to selectedTemplateId + useQuery
for getTemplateById() — ensures full template data (description, attack_procedure,
expected_detection, tool_suggested etc.) loads before modal opens
- DB backfill already applied via SQL: UPDATE audit_logs SET timestamp = NOW()
WHERE timestamp IS NULL (358 rows fixed)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- audit_service: set timestamp=datetime.now(utc) explicitly so DB never stores NULL
- AuditLogPage: formatDate handles null/undefined timestamps (was showing Jan 1 1970)
- nginx.conf: add CSP script-src hash for inline script (sha256-31OgE8E9...)
- system.py: MITRE sync now runs in BackgroundTasks — returns immediately, no more 120s timeout
- mitre_sync_job.py: add _run_data_sources_sync job (every 6h) that checks sync_frequency
and auto-syncs overdue enabled data sources
- SystemPage: MITRE sync result shows "started" vs "complete" message
- test-templates.ts: add updateTemplate() API function
- SystemPage: template name cell is now clickable — opens TemplateDetailModal with
full edit form (name, description, procedure, detection, platform, severity, tool)
and Save / Activate / Deactivate / Close buttons
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Algunos registros de audit_log tienen timestamp=NULL en DB.
AuditLogOut tenia timestamp: datetime (no opcional) causando
ValidationError -> 500 Internal Server Error al listar el audit log.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Permite desactivar la cookie Secure en servidores HTTP via .env.
Por defecto false para la instancia local (192.168.1.93).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- auth: desacopla SECURE_COOKIES de AEGIS_ENV para que el login
funcione sobre HTTP (SECURE_COOKIES=false en servidor local)
- TechniqueCell: button -> Link para href real (right-click, a11y)
- TechniquesPage: añade Link en celda MITRE ID en vista lista
- nginx CSP: amplía connect-src con ws:/wss: para evitar bloqueos
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
kitos