docs(skills): add implement-batch skill for batch issue implementation - #51
Merged
Conversation
GitHub-only, self-contained batch driver customized for okfmem: implement an epic's sub-issues (or an explicit list) onto one branch as a single accumulated commit, one isolated subagent per issue, then a bounded adversarial review pass, finalizing through /open-pr. Adapted from the offlinecv sibling to okfmem's facts: pure-Python stdlib gates (leak gate hard + affected pytest + advisory ruff, no typecheck/build/fallow), ecc:python-reviewer, s-annam/okfmem protected main. Replaces the fixture-PII rule with okfmem's non-negotiable public-repo leak rule, and adds confirmation- discipline-rung and Windows/path invariants as blocking review classes.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Adds the
implement-batchskill — a GitHub-only, self-contained batch driver forokfmem. It implements an epic's sub-issues (or an explicit#,#,#list) ontoone feature branch as a single accumulated commit, running one isolated subagent
per issue (context-isolation is the mechanism, not an optimization), then a
bounded adversarial review pass, finalizing through
/open-pr.Adapted from the
offlinecvsibling and customized to okfmem's facts.No dedicated tracking issue — this is skill tooling. (
Refsomitted.)What changed vs the offlinecv original
scripts/check-leaks.py(hard) +affected
pytest+ advisoryruff check. Dropped the TS/Reacttypecheck/build/
fallowsteps; the full OS-matrixverifyis the PR gate.ecc:python-reviewer(fallbackecc:code-reviewer, then ageneral-purposesubagent running/code-review).no private strings in tracked files;
check-leaks.pymust exit 0.rung (state-changing ops), and Windows/path invariants (pure-ASCII
.ps1,path-by-shape, expected assertion values from the code's own normalizer).
/open-pr(leak-gate preflight + single-commit collapsefor linear history);
s-annam/okfmemprotectedmain, per-issue-keywordauto-close.
Retains: subagent-per-issue isolation,
blocked_bytopological ordering,threaded handoff notes, per-stage self-reported provenance, the single Phase-1
human confirmation gate, and the marker-guarded
## Adversarial reviewPR-bodyappend.
Test plan
python3 scripts/check-leaks.pyexits 0 (scanned 65 files, no leaks)ruff check .reviewed (advisory; no Python changed — docs-only)/implement-batchProvenance
Skill authoring (adapt offlinecv → okfmem) via: Claude Opus 4.8 (1M context) — high effort
Verification: CI
verify— pending