j4k-oss/agent-skills
Watch
1
0
Fork
You've already forked agent-skills
 0
Code Issues 6 Pull requests Releases Packages Activity Actions

fix(skills): keep sub-agent instructions harness-neutral #92

Merged
jercik merged 2 commits from fix/harness-neutral-subagent-skills into feat/share-subagent-model-rules 2026-10-01 06:50:15 +00:00
Conversation 2 Commits 2 Files changed 3 +5 -5
jercik commented 2026-10-01 05:55:47 +00:00
Owner
Copy link

Skills now leave model and effort to the per-harness rules under rules/claude/ and rules/codex/. Three skill sites lose their fallback model and effort wording and the Claude Code-specific "Agent tool" and subagent_type=Explore. Each keeps only what the sub-agent delivers (facts or a conclusion), which is what those rules select on. Without a rule, the harness default picks the model.

In 16 test runs, agents given the Claude rule picked opus at xhigh for conclusions and medium for facts. Agents given the Codex rule picked gpt-6-astra at high and gpt-6-sol at medium.

Two neighboring lines also changed. improve-codebase-architecture now applies the deletion test with the outcomes the skill defines. In review-agent-instructions, the bar on editing a retest no longer ends once the retest passes.

Stacked on #91. Retarget to main after #91 merges.

Skills now leave model and effort to the per-harness rules under `rules/claude/` and `rules/codex/`. Three skill sites lose their fallback model and effort wording and the Claude Code-specific "Agent tool" and `subagent_type=Explore`. Each keeps only what the sub-agent delivers (facts or a conclusion), which is what those rules select on. Without a rule, the harness default picks the model. In 16 test runs, agents given the Claude rule picked `opus` at `xhigh` for conclusions and `medium` for facts. Agents given the Codex rule picked `gpt-6-astra` at `high` and `gpt-6-sol` at `medium`. Two neighboring lines also changed. `improve-codebase-architecture` now applies the deletion test with the outcomes the skill defines. In `review-agent-instructions`, the bar on editing a retest no longer ends once the retest passes. Stacked on #91. Retarget to `main` after #91 merges.
fix(skills): keep sub-agent instructions harness-neutral
All checks were successful
commit-msg / commitlint (pull_request) Successful in 29s
Details
Node tests / node:test (pull_request) Successful in 2m6s
Details
Review / Review (pull_request_target) Successful in 5m6s
Details
466d37bef7
forgejo-actions commented 2026-10-01 06:00:53 +00:00
Copy link

Review 01M3V0WM370ZE9G4QXS4B983WN — head 0492ee1ff904a13c468c52001ecc41bc6540ae98

Review — j4k-oss/agent-skills @ 043722e801

Scope: diff against base tree a9f5aa9039f8
Status: dispatched — coverage complete (3/3 slots terminal)
Facts: current review-wide projection

Computed under:

{
  "abandonment": "abandonment-v1",
  "anchor_recipe": 1,
  "batch_policy": "batch-v1",
  "coverage": "coverage-v3",
  "dispatch_policy": "dispatch-v2",
  "grounder_version": 1,
  "grounding_read_rule": "grounding-read-v1",
  "promotion_policy": "promotion-v1",
  "report": "report-v3",
  "tally": "tally-v1",
  "triage_settle": "triage-settle-v2"
}

Findings (0)

No findings survived.

Reviewed:

  • general-bug (whole/default): no-claims
  • writing-quality (whole/default): no-claims
  • test-trimming (whole/default): no-claims

Other claims

  • grounding-pending (0)
  • ungrounded (0)
  • rejected (0)
  • duplicate-of (0)
  • unadjudicated (0)

Coverage

Coverage pass: 01M3V0WM628CAYBWNTXR1M090S
Accounting: complete
Slot health: healthy

lens part arm unit status runs loss
general-bug whole default no-claims 1 no
writing-quality whole default no-claims 1 no
test-trimming whole default no-claims 1 no
<!-- review:summary --> **Review** `01M3V0WM370ZE9G4QXS4B983WN` — head `0492ee1ff904a13c468c52001ecc41bc6540ae98` # Review — j4k-oss/agent-skills @ 043722e80164 Scope: diff against base tree `a9f5aa9039f8` Status: dispatched — coverage complete (3/3 slots terminal) Facts: current review-wide projection Computed under: ```json { "abandonment": "abandonment-v1", "anchor_recipe": 1, "batch_policy": "batch-v1", "coverage": "coverage-v3", "dispatch_policy": "dispatch-v2", "grounder_version": 1, "grounding_read_rule": "grounding-read-v1", "promotion_policy": "promotion-v1", "report": "report-v3", "tally": "tally-v1", "triage_settle": "triage-settle-v2" } ``` ## Findings (0) No findings survived. Reviewed: - general-bug (whole/default): no-claims - writing-quality (whole/default): no-claims - test-trimming (whole/default): no-claims ## Other claims - grounding-pending (0) - ungrounded (0) - rejected (0) - duplicate-of (0) - unadjudicated (0) ## Coverage Coverage pass: 01M3V0WM628CAYBWNTXR1M090S Accounting: complete Slot health: healthy | lens | part | arm | unit status | runs | loss | | --- | --- | --- | --- | --- | --- | | general-bug | whole | default | no-claims | 1 | no | | writing-quality | whole | default | no-claims | 1 | no | | test-trimming | whole | default | no-claims | 1 | no |
🚀 1
fix(skills): align the deletion-test and retest wording with their definitions
All checks were successful
commit-msg / commitlint (pull_request) Successful in 26s
Details
Node tests / node:test (pull_request) Successful in 1m38s
Details
Review / Review (pull_request_target) Successful in 1m50s
Details
0492ee1ff9
jercik commented 2026-10-01 06:02:05 +00:00
Author
Owner
Copy link

Replying to review comment #98076

Agreed on both claims, fixed in 0492ee1ff9.

  • 01M3V0PX6ADCGEEDJXRSN91N22: the process step now uses the deletion test's defined outcomes. Complexity that vanishes marks a pass-through; complexity that reappears across callers means the module is earning its keep.
  • 01M3V0NDXVCASMG8024KSHF59V: a retest now runs against unchanged expectations and reports its result before any change, so a pass no longer ends the restriction.
> Replying to review comment #98076 Agreed on both claims, fixed in 0492ee1ff904a13c468c52001ecc41bc6540ae98. - `01M3V0PX6ADCGEEDJXRSN91N22`: the process step now uses the deletion test's defined outcomes. Complexity that vanishes marks a pass-through; complexity that reappears across callers means the module is earning its keep. - `01M3V0NDXVCASMG8024KSHF59V`: a retest now runs against unchanged expectations and reports its result before any change, so a pass no longer ends the restriction.
jercik merged commit 1cd99e7d91 into feat/share-subagent-model-rules 2026-10-01 06:50:15 +00:00
jercik deleted branch fix/harness-neutral-subagent-skills 2026-10-01 06:50:15 +00:00
Sign in to join this conversation.
Reviewers
No reviewers
Labels
Clear labels
No items
No labels
Milestone
Clear milestone
No items
No milestone
Assignees
Clear assignees
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
j4k-oss/agent-skills!92
No description provided.