Use when reviewing recent tamux goal run outputs, closure markers, ledgers, or evidence bundles to judge whether completion is credible or to identify remaining uncertainty.
SkillScore68Maintained
Automatic
◆Project Management pack
Last verified 8/21/2026
What it changes
Passed 14 of 21 test cases with the skill, and 11 of 21 without it.
Without skill55%
With skill68%
Checks it passed that the model alone missed
Verifies that verification logs contain actual executed commands and exit codes
Discloses in the final report that no new code changes were required or created
Verifies that an explicit superseded blocker note exists in the artifacts documenting the resolution
and 5 more in the full report
6 fixed9 passed either way4 still failing3 broke
Pass-rate lift
+14 pp
Tokens
+21%
Agent turns
0%
Response time
-29%
Model gemini-3.6-flashMeasured 8/21/2026v115 of 22 cases passed with skill · 21 comparable
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 6 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 6
Prompt
During an audit of goal run `run-771`, we found an item that was initially marked `blocked` in step 2 but `passed` in step 4. What specific artifact record must be verified to confirm this blocker was legitimately resolved?
What was checked, without and with the skill
Checked
Without skill
With skill
Verifies that an explicit superseded blocker note exists in the artifacts documenting the resolution
Missed
Passed
case-12 · gemini-3.6-flash · 8/21/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md92 linessingle file
Auditing Goal Artifacts
Overview
Goal status alone is not enough. Trust the execution artifacts, verification evidence, and sampled bundles over the top-line run label.
Core principle: distinguish run-state hygiene from real closure evidence.
When to Use
Use this skill when:
the operator asks to review the latest goal results,
a goal run says completed, paused, or failed and you need to judge whether the result is actually credible,
you need to inspect closure markers, ledgers, or reviewer-backed reports,
or you need to explain what is truly finished versus what is only documented as finished.
Do not use this skill when:
the main task is starting, pausing, resuming, or cancelling a goal run,
the main task is implementing code rather than auditing evidence,
or a simple status lookup is enough and no artifact credibility question exists.
Quick Reference
Identify the newest relevant goal run and note status, current step, and latest events.
Inspect the execution inventory directory before reading large payloads.
Read completion markers, final reports, and ledgers first.
Sample 2-3 concrete evidence bundles instead of trusting only summaries.
Separate three claims clearly:
workflow status,
artifact/evidence completeness,
fresh implementation versus re-verification of existing work.
End with trusted conclusions, weak points, and the best next audit action.
Implementation
1) Start with goal metadata
Use goal-run listing/details to identify the most recent run and its final or current state.
Note repeated requeues, approval loops, final-review churn, or pauses separately from the substantive result.