chore(ai-review): the gateway review route moved #6

Merged
A-Guevara merged 1 commit from chore/ai-review-20260923 into main 2026-09-23 17:24:28 +00:00
Owner

The AI review gateway renamed its routes (nis/forgejo-stack ADR-0002) and kept no aliases, so the workflow this repo carries now gets a 404 and posts a could-not-review comment. This re-seeds the canonical ai/ai-review.yml, which calls the new route. Merge it to get AI review back on this repository. Advisory only, never a merge gate.

The AI review gateway renamed its routes (nis/forgejo-stack ADR-0002) and kept no aliases, so the workflow this repo carries now gets a 404 and posts a could-not-review comment. This re-seeds the canonical ai/ai-review.yml, which calls the new route. Merge it to get AI review back on this repository. Advisory only, never a merge gate.
chore(ai-review): the gateway review route moved (ADR-0002)
All checks were successful
hello / hello (push) Successful in 11s
ai-review / review (pull_request) Successful in 48s
64aca563ce
ai-review-bot left a comment

AI review · advisory

Verdict: 1 thing worth fixing (1 medium).

Findings that didn't map to a diff line:

tti/coding-standards/.forgejo/workflows/ai-review.yml:101 · MEDIUM — Standard out of sync with implementation
The coding‑standard file still documents the old request body (including "repo" and "pr" fields) and the old endpoint "/v1/review", which no longer matches the workflow’s updated API call, risking confusion and future regressions.

Fix: Update the coding‑standard file to reflect the new request body and endpoint used by the workflow.

⚑ panel: GPT-OSS 120B · Gemma 4 31B · Devstral 2 123B · Laguna S 2.1 — 1 distinct, 1 confirmed, 0 below threshold, 0 refuted · web: not used · context: 1 files under review · 7 codebase · 6 standards chunks

Panel debate — how this review was reached

Grounding — context: 1 files under review · 7 codebase · 6 standards chunks

Round 1 — independent reviews

  • GPT-OSS 120B (1 finding, confidence 0.97 · thought 939 + wrote 747 of 16.0k tokens · 9 s): The workflow changes are consistent, but the coding‑standard documentation is stale and should be updated to match the new API contract.
  • Gemma 4 31B (0 findings, confidence 1.0 · thought 1.6k + wrote 249 of 16.0k tokens · 36 s): The changes correctly migrate the AI review API call from a single endpoint to a RESTful route and update the corresponding transcript link.
  • Devstral 2 123B (0 findings, confidence 1.0 · wrote 42 of 16.0k tokens · 3 s): The changes are correct and consistent with the codebase standards.
  • Laguna S 2.1 (0 findings, confidence 0.95 · thought 7.9k + wrote 238 of 65.5k tokens · 57 s): No defects found. The changed lines correctly migrate the BFF call from the old /v1/review route to the Forgejo-shaped /v1/repos/{o}/{r}/pulls/{n}/reviews route: repo and PR number are moved from the

Round 2 — cross-examination

  • GPT-OSS 120B#1 Standard out of sync with implementation · confirmed: Devstral 2 123B, Laguna S 2.1 · refuted: Gemma 4 31B · support 2

Synthesis — Laguna S 2.1 wrote the final review from 1 confirmed finding · promotion: support ≥ 2, and no refutation at high severity.

Transcript rv-20260923171645-d2b899 — full round outputs, web results, and model reasoning are viewable by anyone with access to this repository via the AI gateway.

Advisory — never a merge gate. Findings are ordered by how well the panel's own evidence checks out, strongest first. React 👍/👎 on any inline comment to tell it whether it was worth flagging — that is the only feedback this system gets, and every threshold in it is tuned from those reactions. Transcript rv-20260923171645-d2b899.

### AI review · advisory <!-- tti-rv:rv-20260923171645-d2b899: --> **Verdict: 1 thing worth fixing** (1 medium). Findings that didn't map to a diff line: **`tti/coding-standards/.forgejo/workflows/ai-review.yml:101`** · MEDIUM — Standard out of sync with implementation The coding‑standard file still documents the old request body (including "repo" and "pr" fields) and the old endpoint "/v1/review", which no longer matches the workflow’s updated API call, risking confusion and future regressions. > **Fix:** Update the coding‑standard file to reflect the new request body and endpoint used by the workflow. <sub>⚑ panel: GPT-OSS 120B · Gemma 4 31B · Devstral 2 123B · Laguna S 2.1 — 1 distinct, 1 confirmed, 0 below threshold, 0 refuted · web: not used · context: 1 files under review · 7 codebase · 6 standards chunks</sub> <details> <summary>Panel debate — how this review was reached</summary> **Grounding** — context: 1 files under review · 7 codebase · 6 standards chunks **Round 1 — independent reviews** - **GPT-OSS 120B** (1 finding, confidence 0.97 · thought 939 + wrote 747 of 16.0k tokens · 9 s): The workflow changes are consistent, but the coding‑standard documentation is stale and should be updated to match the new API contract. - **Gemma 4 31B** (0 findings, confidence 1.0 · thought 1.6k + wrote 249 of 16.0k tokens · 36 s): The changes correctly migrate the AI review API call from a single endpoint to a RESTful route and update the corresponding transcript link. - **Devstral 2 123B** (0 findings, confidence 1.0 · wrote 42 of 16.0k tokens · 3 s): The changes are correct and consistent with the codebase standards. - **Laguna S 2.1** (0 findings, confidence 0.95 · thought 7.9k + wrote 238 of 65.5k tokens · 57 s): No defects found. The changed lines correctly migrate the BFF call from the old /v1/review route to the Forgejo-shaped /v1/repos/{o}/{r}/pulls/{n}/reviews route: repo and PR number are moved from the **Round 2 — cross-examination** - `GPT-OSS 120B#1` Standard out of sync with implementation · confirmed: Devstral 2 123B, Laguna S 2.1 · refuted: Gemma 4 31B · support 2 **Synthesis** — Laguna S 2.1 wrote the final review from 1 confirmed finding · promotion: support ≥ 2, and no refutation at high severity. <sub>Transcript `rv-20260923171645-d2b899` — full round outputs, web results, and model reasoning are viewable by anyone with access to this repository via the AI gateway.</sub> </details> <sub>Advisory — never a merge gate. Findings are ordered by how well the panel's own evidence checks out, strongest first. React 👍/👎 on any inline comment to tell it whether it was worth flagging — that is the only feedback this system gets, and every threshold in it is tuned from those reactions. Transcript `rv-20260923171645-d2b899`.</sub>
A-Guevara deleted branch chore/ai-review-20260923 2026-09-23 17:24:30 +00:00
Sign in to join this conversation.
No reviewers
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
tti/welcome!6
No description provided.