Semih
dfc19c8f13
refactor(integrations): migrate issue tracker GitHub → Gitea
...
- apps/web/src/lib/gitea.ts: createIssue / getIssue / verifyWebhookSignature / buildIssueBody
* Endpoint: git.semih.ai/api/v1 (configurable via GITEA_BASE_URL)
* Auth: 'Authorization: token <PAT>' (Gitea convention)
* Labels: Gitea expects numeric IDs not strings → ensureLabels() resolves/creates
with color coding (P0/P1 red, P2 yellow, P3 green, type-* grey, default blue)
* Webhook signature: X-Gitea-Signature (hex, no sha256= prefix)
- apps/worker/src/lib/gitea.ts: read-only getIssue() for sync polling
- _actions.ts + github-sync.ts now import from /lib/gitea
- Removed old apps/{web,worker}/.../lib/github.ts + /api/webhooks/github route
(the receiver was already dead — sp.semih.ai is Tailscale-only)
- UI: 'GitHub' label → 'Gitea' on insight detail card
- github-sync job filters by githubIssueUrl.startsWith(GITEA_BASE_URL) so legacy
GitHub-hosted insights (semihyesilyurt/sase.tr#20 ) stay frozen rather than
collide with same-numbered Gitea issues at root/sase.tr.
Env migration (Coolify, panel-web + panel-worker):
- removed: GITHUB_TOKEN, GITHUB_REPO_SASE, GITHUB_WEBHOOK_SECRET
- added: GITEA_TOKEN, GITEA_REPO_SASE=root/sase.tr, GITEA_BASE_URL=https://git.semih.ai
Provisioned Gitea PAT 'super-panel-insights' (scopes: write:repository + write:issue),
stored in Bitwarden.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com >
2026-05-14 09:36:57 +00:00
Semih
d0c497d12d
feat(panel-taxonomy): track Faz 2 sase.tr events (page_visibility, idle, shortcuts, referrals, settings, bulk)
...
7 new event names added to TRACKED_EVENTS so enrich.ts fetches them for compression
timeline interleaving + tagger context. No new tag rules yet (Faz 2 is enrichment,
not signal extraction).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com >
2026-05-14 06:28:20 +00:00
Semih
b83d28c119
feat(eval): POST /api/insights/eval-sets thin wrapper (create + optional run)
...
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com >
2026-05-14 06:15:39 +00:00
Semih
b635b529e2
feat(phase6e): eval framework + prompt editor + daily brief + retention + bundle mode
...
Schema:
- eval_sets (promptTag, cases JSON: [{id, timeline, expected: {...}}])
- eval_runs (per-execution scoring: passedSchema/Severity/Rubric, cost, tokens, results)
Prompts:
- 2 new seed templates: upgrade_hesitation (flash), pattern_bundle (pro)
- pickPromptTag() routes 'upgrade_hesitation' tag to upgrade_hesitation prompt
- Editor UI at /insights/settings/prompts/[id]: edit system/user/schema/tier/temp,
publishPromptVersion() creates new version + deactivates old (CRUD with auto bump)
- setPromptActive() to toggle versions
Daily Brief (/insights/brief):
- Last 24h: sessions processed, insights produced (with severity breakdown),
cost, cache hit rate; today's top 5 priorities; week stats (shipped/validated/regressed)
Retention cron (04:15 UTC daily):
- Delete sessions_meta + session_custom_events older than 90d (unless referenced by
active insight)
- Delete compressed_sessions rows + MinIO timeline blobs older than 180d
- raw_metadata 30d (currently no-op; we don't persist raw metadata to MinIO)
Eval framework:
- apps/worker/src/jobs/eval-run.ts: runs cases against current/specified prompt version,
scores schema_pass + severity_match + rubric_substring; stores EvalRun
- apps/worker BullMQ queue handler for 'eval-run' job name
- apps/web installed bullmq; /lib/queue.ts thin Queue accessor
- Web actions: createEvalSet, triggerEvalRun (enqueues job to insight-pipeline queue)
- UI: /insights/settings/eval-sets list, /new create form (paste JSON cases),
/[id] detail with Run button + recent runs + per-case JSON
Bundle mode (analyze job):
- Pull 3x batch, group by fingerprint
- Groups ≥ INSIGHT_BUNDLE_THRESHOLD (default 3) → use 'pattern_bundle' prompt
- Timeline = primary rep + PATTERN BUNDLE summary block (occurrences, unique users, deltas)
- Insight stores ALL group session IDs as relatedSessionIds; all marked analyzed in one tx
- Cost amortized: 1 LLM call per group
Nav/Cmd+K:
- Inbox header links: Brief, Patterns, Eval sets
- Palette: Daily brief, Eval sets entries
Deferred to backlog: embedding-similarity cross-fingerprint clustering, Telegram brief delivery,
Sase.tr-side data-private audit (separate repo).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com >
2026-05-14 06:11:48 +00:00
Semih
9c7682ae63
feat(phase6d): GitHub polling sync — outbound-only state mirror (Tailscale-only panel)
...
Webhook can't reach sp.semih.ai from internet (Tailscale-only DNS). Switched to a
worker job that polls GitHub for tracked issue states every 10 minutes.
apps/worker/src/lib/github.ts: read-only getIssue() client
apps/worker/src/jobs/github-sync.ts: scan insights with githubIssueNumber + not validated/dismissed/duplicate,
fetch remote state, reflect transitions:
open → closed = shipped + shippedAt + validationStartedAt
closed → open = in_progress + clear validation state
Scheduler: github-sync@*/10min added.
The /api/webhooks/github route stays (works if reachable in future) but is no longer
the source of truth.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com >
2026-05-14 05:54:58 +00:00
Semih
0c6350db2d
feat(insights): POST /api/insights/[id]/create-issue thin wrapper around server action
...
Useful for programmatic smoke tests and future automation. Auth-gated via underlying action.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com >
2026-05-14 05:49:20 +00:00
Semih
c5acdfc8ec
feat(phase6d): GitHub action loop — issue creation, webhook, validation cron, patterns view
...
Schema:
- Insight.+githubIssueNumber (user-visible #N, separate from id BigInt)
GitHub integration (apps/web/src/lib/github.ts):
- repoFor(projectKey): env-based GITHUB_REPO_<KEY>=owner/repo mapping
- createIssue / getIssue REST wrappers
- verifyWebhookSignature (HMAC-SHA256 timing-safe)
- buildIssueBody: renders structured markdown from insight + LLM body
(hypothesis, reproduce steps, affected route/provider, quick/long fixes,
suggested investigation, evidence links to panel, DoD checklist)
Server action createGithubIssueForInsight:
- Auth-gated, audited; idempotent (refuses if issue already exists)
- Labels: insight-driven, severity-<P>, type-<T>, <project>-pilot
- Sets status=in_backlog, stores githubIssueUrl/Id/Number/State
Webhook /api/webhooks/github:
- Signature verify with GITHUB_WEBHOOK_SECRET
- issues.closed → status=shipped + shippedAt + validationStartedAt
- issues.reopened → status=in_progress + clear validation state
- issues.opened → status=in_backlog
- PR linking placeholder (passthrough only for now)
Validation cron (worker, daily 5:00 UTC):
- For each insight in 'shipped' state:
- Count sessions with same fingerprint after shippedAt
- >= INSIGHT_REGRESSION_THRESHOLD (default 3) → status=regressed + regressionDetected=true
- validationPeriodDays elapsed with no regression → status=validated + validatedAt
UI:
- Insight detail: GithubActions card — Create button (when no issue),
external link + issue # + state (when present)
- New /insights/patterns page: clusters insights by type + affected_route/provider,
shows ≥2-insight or ≥5-occurrence groups sorted by max severity
- Inbox header link to Patterns
- Cmd+K palette: Patterns entry
Env needed: GITHUB_TOKEN, GITHUB_REPO_SASE, GITHUB_WEBHOOK_SECRET.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com >
2026-05-14 05:45:48 +00:00
Semih
5beb89f771
feat(phase6c): insight inbox actions + filters + keyboard shortcuts + budget UI + prompt registry view
...
Server actions (apps/web/src/app/insights/_actions.ts):
- setStatus / bulkSetStatus (11 allowed states)
- setFounderNotes / setFounderPriority / setSeverityOverride
- updateBudgetSetting (key whitelist)
All audit-logged + revalidate /insights paths.
Inbox enhancements:
- Filter bar: severity / status (active default = new+triaged+in_backlog+in_progress+regressed)
/ type chips with live counts, query-param driven
- Per-row action buttons: Backlog / Defer / Dismiss with toast feedback
- Keyboard shortcuts (j/k navigate, Enter open, b backlog, e defer, d dismiss, ? help)
Skipped when input/textarea focused.
Insight detail (/insights/i/[id]):
- FounderForm component: 6 status buttons, severity override dropdown,
founder_priority pin input, founder_notes textarea (saved separately)
Budget settings (/insights/settings/budgets):
- Editable rows for monthly/daily caps, per-call max, min_score_for_analysis,
cache_ttl_hours, analysis_paused (kill switch)
Prompt registry (read-only, full editor in 6e):
- /insights/settings/prompts list
- /insights/settings/prompts/[id] detail w/ system prompt, user template,
output schema, recent cost ledger rows for this template
Cmd+K palette extended with Insights group:
- Inbox, Cost dashboard, Pipeline, Budget settings, Prompt registry
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com >
2026-05-13 23:32:41 +00:00
Semih
95c7e8e321
feat(phase6b): LLM analysis layer — DeepSeek + insights + cost ledger + budget guard
...
Schema:
- insights (fingerprint dedup, occurrence_count, related_session_ids[], priority_score)
- cost_ledger (per-call, cache_hit/miss/output token split + USD)
- prompt_templates (versioned, per-tag, with JSON schema + tier + temperature)
- budget_settings (global key/value: monthly_hard_cap_usd, daily_soft/hard, per_call_max, analysis_paused)
Worker:
- lib/deepseek.ts: Anthropic-compat endpoint client (deepseek-v4-flash | deepseek-v4-pro),
usage→USD with V4 promo pricing, cache_read_input_tokens awareness, extractJson() helper
- lib/budget.ts: checkBudget() returns active|soft_throttled|hard_paused|monthly_paused,
forces flash tier on soft cap, halts on hard/monthly cap or analysis_paused
- lib/prompts.ts: 5 seed templates (bug_triage P, ux_friction F, payment_issue P,
onboarding_stuck F, provider_quality F) with embedded JSON schemas + Sase.tr context;
pickPromptTag() maps session tags → prompt
- lib/json-validate.ts: lightweight schema validator (no ajv dep)
- lib/seed-runtime.ts: idempotent upsert of prompts + default budget settings on worker boot
- lib/minio.ts: +getText() for compressed timeline fetch
- jobs/analyze.ts: budget guard → 6h fingerprint cache (attach session to existing insight) →
template lookup → severity-based tier override → DeepSeek call → JSON parse + validate →
insert insight (or aggregate occurrence) → write cost_ledger
- scheduler: analyze@*/4min on insight-pipeline queue
UI:
- /insights (was pipeline view) → now Insight Inbox: priority-sorted list w/ KPI strip
(new/in_backlog/shipped/today $/month $), severity badges, link to detail
- /insights/i/[id]: insight detail with structured body render, related sessions,
per-session cost breakdown, raw JSON collapsible
- /insights/costs: KPI cards (today, month, avg, cache hit), daily 30d bar table,
by-model + by-prompt breakdowns, top 10 expensive, recent errors
- /insights/pipeline: moved old session-pipeline view here
- /insights/sessions/[id]: unchanged session timeline viewer
Defaults:
- monthly cap $30, daily soft $1.50 / hard $3, per-call $0.20, analysis_paused=false
- Severity→Tier: P0/P1=pro, P2/P3/INFO=flash; budget soft-cap forces flash
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com >
2026-05-13 23:01:55 +00:00
Semih
978b71d818
fix(phase6a-v1.2): PostHog event filter client-side + backfill distinct_id + power_user_path rrweb fallback
...
- listSessionEvents: PostHog only honors single event= param; switched to client-side filter
- tag-sessions: backfill posthogDistinctId from getRecording when null (for rows pre-column)
- tagger: power_user_path v1 fallback (auth + clicks>=20 + no errors), vin_decode_no_outcome P3
- ingest upsert: always update posthogDistinctId (idempotent)
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com >
2026-05-13 22:43:44 +00:00
Semih
7606e1b0bf
feat(insights): /api/insights/reprocess auth-gated endpoint to re-run pipeline on existing sessions
...
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com >
2026-05-13 22:26:13 +00:00
Semih
fc6a3f7a39
feat(phase6a-v1.2): dual-mode custom-event enrichment (v1.0 + v2.0 aliases)
...
- Prisma: posthog_person_cache (24h TTL), posthog_group_cache, session_custom_events
- SessionMeta: +posthogDistinctId, +groupKey, +customEventCount
- PostHog client: listSessionEvents, getPerson, getGroup
- posthog-cache.ts: cache-with-stale-refresh for person + group properties
- event-taxonomy.ts: dual-mode alias map (vin_decoded↔vin_decode_initiated,
vin_decode_success↔vin_decode_succeeded, vin_decode_error↔vin_decode_failed,
payment_success↔payment_succeeded). 56 tracked event names total.
- enrich.ts: per-session custom events fetch, persist to session_custom_events,
attach user_properties + group_properties via cache.
- tagger.ts: 14 new custom-event tags
(vin_decode_fail_pattern, provider_reliability_issue, payment_friction,
payment_ui_silent_failure (P0), payment_failed_session, checkout_abandonment,
downgrade_pending, compatibility_quality_gap, parts_export_abandoned,
search_validation_friction, onboarding_stuck, upgrade_hesitation,
power_user_path, at_risk_active_session, provider_mismatch,
api_key_friction, webhook_setup_struggle); group-context priority bump.
- heuristic.ts: customEventPromoteReasons() — 10 custom-event promote signals.
- compress.ts: merge_rrweb_and_custom_events interleaving with ⤷ EVENT: prefix,
±1.5s grouping, standalone flush for events with no nearby rrweb. Adds
company_context header line; key_events lists top-5 custom event counts.
- /insights page: +Events column. /insights/sessions/[id]: collapsible custom
events table + group key display.
Graceful: when Sase.tr emits 0 custom events the pipeline falls back to
rrweb-only tagging from Phase 6a v1.1.
PRD: behavioral-insight-pipeline-prd-1.md (v1.2),
sase-posthog-events-prd.md (Sase.tr-side, separate codebase).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com >
2026-05-13 22:17:18 +00:00
Semih
b9709ed07e
feat(phase3): worker container + event bus + scheduled jobs
...
apps/worker:
- BullMQ nightly scheduler (cron 0 3 * * *)
- Redis Streams consumer-group per wired/active project
- Persists events to Event model
schema:
- Event model (streamId unique, project + type indexed)
2026-05-13 11:03:58 +00:00