Repository navigation
Add runnable checks to the five product skills - #1958
Merged
Merged
Conversation
growth-analytics, business-ops, product-output-safety, positioning and product-discovery shipped prose and evals only. Each now ships one deterministic script that checks a JSON plan, readout or draft against rules quoted from its own SKILL.md, fails closed on absent fields, and exits SKIP/FAIL 1/ERROR 2/INCONCLUSIVE 3: - growth-analytics: scripts/readout_check.py - business-ops: scripts/bizops_check.py - product-output-safety: scripts/output_safety_check.py - positioning: scripts/positioning_check.py - product-discovery: scripts/discovery_check.py Each SKILL.md gains one routing line; the generated INDEX files list the scripts; docs/capability-matrix.md rows name them. Each lane passed an independent review (final verdicts 🔴0 🟡0, at most 🔵3 queued as nits). Combined land: the five lane commits are squashed onto one base to save four CI cycles. Table conflicts (size-budgets.tsv, INDEX.md, capability-matrix.md) were resolved by union, then deduplicated; the INDEX files were regenerated. The slim fast-path size caps for business-ops (6350→6454) and growth-analytics (6000→6011) rise by the routing line. size-budget-raise: .claude/skills/INDEX.md 12344→13037 generated index lists the five new product-skill check scripts (combined land) Tests: 166/166 (readout 17, bizops 23, output-safety 29, positioning 46, discovery 35, slim bizops 9, slim GA 7); ci-tests.d check scripts 6/6 OK Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Five product skills (growth-analytics, business-ops, product-output-safety, positioning, product-discovery) shipped prose and evals with no runnable check. Rival skill packs ship scripts. This PR gives each skill one deterministic checker that enforces only rules quoted from its own SKILL.md:
readout_check.pybizops_check.pyoutput_safety_check.pypositioning_check.pydiscovery_check.pyAll five fail closed on absent fields (FAIL, never a sibling-carried PASS) and use exit codes SKIP / FAIL 1 / ERROR 2 / INCONCLUSIVE 3. Lone-surrogate input returns ERROR rc 2 where tested.
Cost
Each SKILL.md grows by one routing line (about 110–150 bytes). The scripts load only when run.
Review
Each lane commit passed an independent review; final verdicts were 🔴0 🟡0, with at most three 🔵 nits that are queued for the next nits PR. Discovery's first review found 🔴3 🟡4 (fail-open bools, ICE SKIP, surrogate traceback), and positioning's found one 🟡 (a regex bypass); both were fixed before the delta reviews.
Land notes
This branch squashes the five lane commits onto main
31419323d5860b34faa9b39dafe15e632ee34c94as one commit, which saves four CI cycles. The table conflicts insize-budgets.tsv,INDEX.mdandcapability-matrix.mdwere resolved by union and then deduplicated, and the INDEX files were regenerated. The slim fast-path caps for business-ops (6350→6454) and growth-analytics (6000→6011) rise by the routing line.Tests
🤖 Generated with Claude Code