Create release observability and incident-response dashboard #41

New Issue

timmy · 2026-08-19T01:59:02Z

timmy commented

2026-08-19 01:59:02 +00:00

Parent epic: #7

Outcome

Give operators actionable health without surveilling users.

Acceptance criteria

Dashboard covers app/API/queue/model and privacy-safe failure classes.
Alerts have owners, thresholds, and runbook links.
A simulated worker outage produces one actionable alert and graceful manual fallback.

Dependencies

#19 — Instrument privacy-safe model health, latency, abstention, and failure metrics

Required evidence

Test or executable receipt attached to the issue/PR
Privacy/safety boundary checked where applicable
Clean checkout and relevant full suite pass

Delivery

Open a focused branch and PR with Closes #ISSUE. Do not include real medical images in issue comments or Git history.

Parent epic: #7 ## Outcome Give operators actionable health without surveilling users. ## Acceptance criteria - [ ] Dashboard covers app/API/queue/model and privacy-safe failure classes. - [ ] Alerts have owners, thresholds, and runbook links. - [ ] A simulated worker outage produces one actionable alert and graceful manual fallback. ## Dependencies - #19 — Instrument privacy-safe model health, latency, abstention, and failure metrics ## Required evidence - Test or executable receipt attached to the issue/PR - Privacy/safety boundary checked where applicable - Clean checkout and relevant full suite pass ## Delivery Open a focused branch and PR with `Closes #ISSUE`. Do not include real medical images in issue comments or Git history.

timmy added this to the M4 — Production Readiness & Release milestone 2026-08-19 01:59:02 +00:00

timmy added the

type/ops

priority/P1

area/release

labels 2026-08-19 01:59:02 +00:00

timmy referenced this issue

2026-08-19 01:59:03 +00:00

EPIC: Production operations, private beta & release #7

rockachopa referenced this issue from a commit

2026-08-22 20:33:45 +00:00

feat(ops): release observability dashboard and incident-response flow

~~rockachopa referenced this issue 2026-08-22 20:34:13 +00:00~~

Release observability dashboard and incident-response flow (#41) #60

rockachopa referenced a pull request that will close this issue

2026-08-22 20:34:28 +00:00

Release observability dashboard and incident-response flow (#41) #60

rockachopa commented

2026-08-22 20:40:46 +00:00

Implemented in PR #60 — #60 (branch timmy/41-release-observability-dashboard, commit ad8c8a9).

Receipts:

Local operator dashboard over sanitized evidence v1 (app/api/queue/model boundaries, failure classes, bounded counters) — node scripts/release_dashboard.mjs --evidence <file>.
Alert inventory with owner/threshold/severity/runbook anchor per rule (worker.outage, queue.backlog, vision.degraded, app.unhealthy) in docs/RELEASE-OBSERVABILITY.md.
Simulated worker-outage drill executed end-to-end on a loopback fixture only: pre-drill 0 alerts → exactly one PAGE worker.outage alert naming owner release-operator, threshold 1, runbook link; manual fallback AVAILABLE (local journal); DRILL PASS exit 0.
Privacy boundary enforced mechanically by sanitizeEvidence(): session IDs, cookies/tokens, environment dumps, photo payloads/base64, image hashes, note text, emails are dropped or rejected without echo; hostile-payload tests included. No photos, stool records, session identifiers, or secrets in code, tests, docs, or history.
Gates: npm test 80/80 · staging-deploy tests 20/20 · sleek browser acceptance on this tree PASS · npm run check:syntax PASS · npm audit --audit-level=high 0 vulnerabilities · check:diff clean · secret/privacy scan across changed files PASS.
CI: Quality gates / quality (pull_request) success on commit ad8c8a9 (1m44s).

No live host was contacted or modified; nothing merged or deployed.

Implemented in PR #60 — https://forge.alexanderwhitestone.com/git/stackchain/timmy-talking-turd/pulls/60 (branch `timmy/41-release-observability-dashboard`, commit ad8c8a9). Receipts: - Local operator dashboard over sanitized evidence v1 (app/api/queue/model boundaries, failure classes, bounded counters) — `node scripts/release_dashboard.mjs --evidence <file>`. - Alert inventory with owner/threshold/severity/runbook anchor per rule (`worker.outage`, `queue.backlog`, `vision.degraded`, `app.unhealthy`) in docs/RELEASE-OBSERVABILITY.md. - Simulated worker-outage drill executed end-to-end on a loopback fixture only: pre-drill 0 alerts → exactly one PAGE worker.outage alert naming owner release-operator, threshold 1, runbook link; manual fallback AVAILABLE (local journal); DRILL PASS exit 0. - Privacy boundary enforced mechanically by sanitizeEvidence(): session IDs, cookies/tokens, environment dumps, photo payloads/base64, image hashes, note text, emails are dropped or rejected without echo; hostile-payload tests included. No photos, stool records, session identifiers, or secrets in code, tests, docs, or history. - Gates: npm test 80/80 · staging-deploy tests 20/20 · sleek browser acceptance on this tree PASS · npm run check:syntax PASS · npm audit --audit-level=high 0 vulnerabilities · check:diff clean · secret/privacy scan across changed files PASS. - CI: Quality gates / quality (pull_request) success on commit ad8c8a9 (1m44s). No live host was contacted or modified; nothing merged or deployed.

Sign in to join this conversation.

No Milestone

No project

No Assignees

2 Participants

Notifications

Due Date

The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: stackchain/timmy-talking-turd#41