state-hub/docs/evidence/workstream-caller-status-20260711.md
tegwick 34016ee9ac
All checks were successful
CI Smoke / host-smoke (push) Successful in 0s
CI Smoke / container-smoke (push) Successful in 1s
docs: workstream caller status evidence; STATE-WP-0070 T04 progress
2026-07-12 00:21:06 +02:00

40 lines
No EOL
1.6 KiB
Markdown

# Workstream legacy caller status — 2026-07-11
Parent: `STATE-WP-0070` T03/T04
## Meter snapshot
| Window | `GET /workstreams/{workstream_id}` | `POST /progress/ workstream_id` |
| --- | ---: | ---: |
| 7-day (weekly review) | 10,335 (`unknown`) | 14 (`unknown`) |
| 8-hour (live) | **0** | **0** |
Weekly volume is **historical** (pre-2026-07-08 fix-consistency and dashboard
migrations). No legacy REST keys show traffic in the current 8h window.
## Confirmed historical callers (remediated in tree)
1. **`consistency_check.py`** — was `GET /workstreams/{id}` per workplan; now `/workplans/`
2. **Dashboard JS** — was `apiFetch("/workstreams/")`; src migrated; **`dist/` rebuilt 2026-07-11**
3. **`validate_repo_adr.py`**, **`custodian_cli.py`** — migrated to `/workplans/`
4. **`activity-core` resolver** — migrated to `/workplans/index`
5. **`the-custodian/runtime`** — triage prompt + `actions.py` now emit `workplan_id` (2026-07-11)
## Why attribution was `unknown`
Dashboard browser `fetch()` does not send `X-StateHub-Component`. Remote scripts
before identity-header deploy also lacked headers.
## Remaining gate
Seven consecutive **8-hour** zero windows per legacy key before MCP alias
removal and T05 closeout. MCP metering trustworthy after dev-hub MCP restart
(2026-07-10) and pytest guard (`STATEHUB_MCP_LEGACY_METER=off`).
## Recommended follow-up
- Remote SHA audit: confirm all fix-consistency hosts ≥ post-`STATE-WP-0072`
- Add optional `X-StateHub-Component: state-hub.dashboard` to `apiFetch` for
future regressions
- Retry `make -C railiance-platform forgejo-backup` when railiance01 websocket
is stable (unrelated to REST callers)