Executive Summary
Over the 180-day window (2025-12-18 to 2026-06-16), 3 of 7 qualifying PR outcomes were accepted, delivering 205 total objective value. The highest-impact work was two bug fixes (PR #37863, PR #37565) resolving critical safe-outputs and multi-repo checkout failures, contributing 170 of 205 value. All accepted outcomes are attributed to PR Description Updater (0.32 value/AIC). Repository-wide Impact Efficiency is 0.005 — 43,055 AIC was spent mostly on analytics/reporting workflows with no linked-issue PR outcomes; 993 of 1,000 sampled PRs were excluded for lacking Closes #N.
Summary
| Metric |
Value |
| PRs analyzed with linked issue |
merged: 3, closed: 4, open excluded: 0 |
| PRs excluded (no linked closing issue) |
993 (497 merged + 496 closed) |
| Safe output outcomes analyzed |
0 |
| Outcomes mapped to objectives |
7 |
| Unmapped outcomes |
0 |
| Accepted outcome count |
3 |
| Total outcome value |
205 |
| AI Credits |
43,055.27 |
| Impact Efficiency |
0.005 value units / AIC |
Agentic Work by Objective
bug (value: 85) — Highest-impact
| Accepted |
Total Value |
AIC (approx.) |
Efficiency (approx.) |
| 2 |
170 |
~181 |
~0.94 |
- PR #37863 -> #37835: Fix push_to_pull_request_branch branch derivation
- PR #37565 -> #37545: Fix multi-repo checkout when Configure Git credentials clobbers origin
improvement (value: 35)
| Accepted |
Total Value |
AIC (approx.) |
Efficiency (approx.) |
| 1 |
35 |
~90 |
~0.39 |
Highest AIC category (no value): Analytics/reporting stack consumed ~18,502 AIC (43% of total) — daily-experiment-report (3,625), Impact Efficiency Report (2,647), Contribution Check (2,278), Daily Semgrep Scan (2,209), Daily Firewall Logs (2,144) — with zero linked-issue PR outcomes.
Which Workflows Drove That Impact
PR Description Updater — Most value (100% share)
| Outcomes |
Value |
AIC |
Efficiency |
| 3 accepted |
205 |
633.26 |
0.32 |
Top categories: bug (170), improvement (35). Attribution based on gh-aw-agentic-workflow body metadata; creating workflow may differ (workflow-logs.json is empty).
Daily Documentation Healer — 0 value
3 analyzed outcomes (all rejected): PRs #36767, #36371, #35330. AIC: unknown (not in aic-by-workflow.json).
Daily Documentation Updater — 0 value
1 analyzed outcome (rejected): PR #35906. AIC: 323.05.
Top AIC consumers with zero attributed outcomes
| Workflow |
AIC |
Runs |
| daily-experiment-report |
3,625.33 |
2 |
| Impact Efficiency Report |
2,647.42 |
3 |
| Contribution Check |
2,278.16 |
5 |
| Daily Semgrep Scan |
2,208.89 |
2 |
| Daily Firewall Logs Collector and Reporter |
2,143.74 |
1 |
| Glossary Maintainer |
2,039.17 |
2 |
| Daily CLI Tools Exploratory Tester |
1,792.41 |
2 |
| Package Specification Extractor |
1,766.81 |
2 |
Top Outcomes by Outcome Value
Unmapped Outcomes
None. All 7 outcomes mapped via objective-mapping.json. Labels cookie, deep-report, linting, go, go-linters, closed:unknown are absent from the mapping and ignored.
Interpretation
- Impact Efficiency (0.005) vs accepted count (3): count alone hides importance. Bug fixes (85 pts each) are ~2.4x more valuable than the linter improvement (35 pts).
- Bug category is most cost-efficient (~0.94 value/AIC) — highest-value, lowest estimated cost.
- PR Description Updater is the only workflow with attributed delivered value (0.32 efficiency). Attribution caveat: it is a description-update workflow; the underlying code authors may be different.
- Daily Documentation Healer: 3 closed-without-merge documentation PRs suggest a systematic acceptance issue. AIC unknown.
- Top 8 AIC workflows (43%) have zero linked-issue outcomes; value likely delivered through reports and non-PR channels not covered by this metric.
Data Quality
| Gap |
Notes |
| PR linked-issue coverage |
99.3% excluded (993/1000). Leading metric gap. |
| Safe-output outcomes |
evaluations.jsonl is empty; 0 outcomes. |
| Workflow attribution |
workflow-logs.json is empty; attribution from body markers only. Merged PR creating workflows unknown. |
| Daily Doc Healer AIC |
Not in aic-by-workflow.json; impact efficiency not computable. |
| Per-category AIC |
Approximated: 633.26 total / 7 runs = ~90.5 AIC/run for PR Description Updater. |
AIC source: Deterministic — aic-by-workflow.json (48 snapshots, token-audit-memory). Total: 43,055.27. No live fallback used.
Human Work
369 merged PRs identified as potentially human-authored (no gh-aw-agentic-workflow body marker). All 3 PRs with linked issues had agentic markers, so zero human PRs are traceable to objectives.
| Metric |
Value |
| Potentially human-authored merged PRs |
369 |
| With linked closing issue |
0 |
| Mapped to objective |
0 |
No AIC computed for human work. Caveat: empty workflow-logs.json means human/agentic distinction relies solely on body markers and may inflate the human count.
Impact Efficiency Report workflow · Run #27643083066 · 2026-06-16 · Window: 2025-12-18 to 2026-06-16 (180 days) · Replaces #39612
Generated by 📊 Impact Efficiency Report · ◷
Executive Summary
Over the 180-day window (2025-12-18 to 2026-06-16), 3 of 7 qualifying PR outcomes were accepted, delivering 205 total objective value. The highest-impact work was two bug fixes (PR #37863, PR #37565) resolving critical safe-outputs and multi-repo checkout failures, contributing 170 of 205 value. All accepted outcomes are attributed to PR Description Updater (0.32 value/AIC). Repository-wide Impact Efficiency is 0.005 — 43,055 AIC was spent mostly on analytics/reporting workflows with no linked-issue PR outcomes; 993 of 1,000 sampled PRs were excluded for lacking
Closes #N.Summary
Agentic Work by Objective
bug (value: 85) — Highest-impact
improvement (value: 35)
Highest AIC category (no value): Analytics/reporting stack consumed ~18,502 AIC (43% of total) — daily-experiment-report (3,625), Impact Efficiency Report (2,647), Contribution Check (2,278), Daily Semgrep Scan (2,209), Daily Firewall Logs (2,144) — with zero linked-issue PR outcomes.
Which Workflows Drove That Impact
PR Description Updater — Most value (100% share)
Top categories: bug (170), improvement (35). Attribution based on
gh-aw-agentic-workflowbody metadata; creating workflow may differ (workflow-logs.json is empty).Daily Documentation Healer — 0 value
3 analyzed outcomes (all rejected): PRs #36767, #36371, #35330. AIC: unknown (not in aic-by-workflow.json).
Daily Documentation Updater — 0 value
1 analyzed outcome (rejected): PR #35906. AIC: 323.05.
Top AIC consumers with zero attributed outcomes
Top Outcomes by Outcome Value
Unmapped Outcomes
None. All 7 outcomes mapped via objective-mapping.json. Labels
cookie,deep-report,linting,go,go-linters,closed:unknownare absent from the mapping and ignored.Interpretation
Data Quality
AIC source: Deterministic — aic-by-workflow.json (48 snapshots, token-audit-memory). Total: 43,055.27. No live fallback used.
Human Work
369 merged PRs identified as potentially human-authored (no
gh-aw-agentic-workflowbody marker). All 3 PRs with linked issues had agentic markers, so zero human PRs are traceable to objectives.No AIC computed for human work. Caveat: empty workflow-logs.json means human/agentic distinction relies solely on body markers and may inflate the human count.
Impact Efficiency Report workflow · Run #27643083066 · 2026-06-16 · Window: 2025-12-18 to 2026-06-16 (180 days) · Replaces #39612