docs(#5925): fix stale OTEL scoped-forwarding example in workflow-contracts.md

[fullsend-ai-coder]

Summary

Fixes a factually incorrect example in the normative "When reviewing PRs" section of docs/contributing/workflow-contracts.md. The parenthetical claimed OTEL_EXPORTER_OTLP_TRACES_HEADERS is "deliberately forwarded only to the triage job" — in reality, the secret is explicitly forwarded to every inline stage job in reusable-dispatch.yml (triage, code, review, fix, retro, prioritize, harness-run).

Changes

• Updated the OTEL forwarding example to accurately state that the secret is forwarded to every inline stage job because each stage runs an agent that emits traces
• Removed the misleading "scoped" qualifier from "explicit, scoped forwarding" since this particular secret is not an example of scoped forwarding

Testing

• make lint-md-links passes — no broken links introduced
• Verified the corrected description matches the actual env: blocks in reusable-dispatch.yml (OTEL secrets appear in all 7 inline stage jobs)
• No test changes needed — TestOTELHeadersSecretThreading validates the actual secret threading in the workflow file, not the documentation text

Checklist

• PR title follows Conventional Commits (correct type, ! for breaking changes)
• Documentation-only change — no code or behavioral impact

---

Closes #5925

Post-script verification

• Branch is not main/master (agent/5925-fix-stale-otel-example)
• Secret scan passed (gitleaks — 19f058a97146a1b014192195afe41d247525e0f6..HEAD)
• PR body secret scan passed (gitleaks — no-git)

[fullsend-ai-review]

🤖 Finished Review · ✅ Success · Started 10:01 AM UTC · Completed 10:11 AM UTC
Commit: 35e0534 · View workflow run →

[github-actions]

Site preview

Preview: https://4bf28d63-site.fullsend-ai.workers.dev

Commit: 35e0534ace52a53769b9909793f67c1f82edb8e7

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[fullsend-ai-review]

Looks good to me

---

Labels: PR fixes documentation under docs/contributing/

[fullsend-ai-retro]

🤖 Finished Retro · ✅ Success · Started 10:42 AM UTC · Completed 10:53 AM UTC
Commit: 35e0534 · View workflow run →

[fullsend-ai-retro]

Retro: PR #5926 — docs fix for stale OTEL forwarding example

Timeline

This PR was the product of a fully autonomous retro→triage→code→review pipeline. The retro agent on PR #5887 identified that the review agent had correctly flagged a factually incorrect example in docs/contributing/workflow-contracts.md (claiming OTEL secrets are scoped to the triage job when they are forwarded to all 7 inline stage jobs), but the author merged #5887 without addressing the finding. The retro agent filed issue #5925 with a well-documented description.

The full pipeline executed cleanly:

• 09:50 — Issue Fix stale scoped-forwarding example in workflow-contracts.md #5925 filed by retro agent
• 09:50–09:54 — Triage confirmed the issue and labeled ready-to-code (~4 min)
• 09:54–10:00 — Code agent produced a single-line fix (~6 min)
• 10:00 — PR docs(#5925): fix stale OTEL scoped-forwarding example in workflow-contracts.md #5926 opened
• 10:00–10:11 — Review agent approved with LGTM (~11 min)
• 10:34 — Human reviewer (rh-hemartin) approved without comments
• 10:40 — Merged via merge queue

Total: ~50 minutes, zero rework, single commit.

Assessment

This workflow went well. Key observations:

1. Retro→code pipeline success story. The retro agent correctly escalated a valid review finding that was dismissed on the original PR. The code agent scoped the fix precisely (1 line changed in 1 file). The review agent correctly approved a straightforward factual correction. This is the system working as designed — review findings that get overridden by authors are surfaced as issues and resolved autonomously.

2. Duplicate dispatches handled gracefully. GitHub delivered duplicate webhooks for both the issue creation and PR open events, resulting in extra fullsend.yaml runs. The downstream cancel-in-progress: true concurrency group on review.yml cancelled the duplicate review run within 1 second. This is extensively tracked by existing issues (feat: dispatch-layer dedup and per-issue concurrency guard to prevent duplicate agent runs #766, Deduplicate review dispatches for the same HEAD SHA #1452, Prioritize same-commit review dedup — PR #944 shows 10x waste on a single SHA #5139, Consolidate 22+ overlapping "skip retro/agent dispatch" issues into a single design issue #5817).

3. Docs-only autonomy evidence. Both the agent and human approved this single-file docs correction without inline comments. The ~23 minute gap between agent approval (10:11) and human approval (10:34) was the primary latency contributor. This is additional evidence for existing tracking issues: agents#427 (docs-only PR competence), #4765 and #5289 (docs-only autonomy readiness). Per #5156, zero-comment human approval is not independently strong autonomy evidence, but it is consistent with the existing body of evidence for this change class.

Proposals

No new proposals. All potential improvements identified are already covered by existing open issues with active tracking.