chore: close READINESS-AUDIT S2–S11 (skill + agents docs)#67
Conversation
- S2 drop Bash(npx tsx:*) from allowed-tools (security: tsx ran any TS file)
- S3 rewrite description against 17:319-328 launch template
- S4 add 5-step "DO NOT write prose. Instead:" procedural block
- S5 add ## Trigger phrases section
- S6 add ## Examples (prompt → YAML → URL) using bundled examples/*.yaml
- S7 prepend `openhop --version` not-found check before health probe
- S8 AGENTS.md: drop seven non-existent rules/*.md promises, fix dead
link at line 103 (now points at PATCH Operations section in
SKILL.md)
- S9 drop stale "Prefix all commands with the repo path" sentence
- S10 fold into S4 step 4 (push --json → parse → return url)
- S11 fold into S4: rewrite the bare-localhost guidance to point agents
at the per-flow URL printed by push
Verified: shared 93/93 pass (skill-md-sync 3/3); SKILL.md frontmatter
parses (yaml.parse → name/allowed-tools/description all readable).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (1)
🚧 Files skipped from review as they are similar to previous changes (1)
WalkthroughUpdated OpenHop documentation: AGENTS.md now points agents to ChangesOpenHop Documentation Consolidation
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes Possibly related PRs
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Review rate limit: 9/10 reviews remaining, refill in 6 minutes. Comment |
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@skills/openhop/SKILL.md`:
- Line 45: Update the phrase "brand new flows" in the SKILL.md content to use
the compound adjective hyphenation: change it to "brand-new flows" so the
modifier is correctly hyphenated before the noun; search for the exact string
"brand new flows" in the file and replace it with "brand-new flows".
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro Plus
Run ID: 5e7b7537-e62b-4460-b023-acfcc3a1ac0b
📒 Files selected for processing (2)
AGENTS.mdskills/openhop/SKILL.md
Trailing whitespace alignment on the new ## Examples table — content unchanged. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
CodeRabbit nit on PR #67 (LanguageTool QB_NEW_EN_HYPHEN). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
There was a problem hiding this comment.
♻️ Duplicate comments (1)
skills/openhop/SKILL.md (1)
45-45:⚠️ Potential issue | 🟡 Minor | ⚡ Quick winFix hyphenation.
"brand new" should be hyphenated when used as a compound adjective before a noun.
✏️ Proposed fix
-For brand new flows, sketch your own YAML against the Schema Reference below and push the same way. +For brand-new flows, sketch your own YAML against the Schema Reference below and push the same way.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@skills/openhop/SKILL.md` at line 45, Change the phrase "For brand new flows, sketch your own YAML against the Schema Reference below and push the same way." to use the hyphenated compound adjective "brand-new" (i.e., "For brand-new flows, ...") so the sentence correctly hyphenates the compound adjective before the noun; update the SKILL.md sentence matching that exact text.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Duplicate comments:
In `@skills/openhop/SKILL.md`:
- Line 45: Change the phrase "For brand new flows, sketch your own YAML against
the Schema Reference below and push the same way." to use the hyphenated
compound adjective "brand-new" (i.e., "For brand-new flows, ...") so the
sentence correctly hyphenates the compound adjective before the noun; update the
SKILL.md sentence matching that exact text.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro Plus
Run ID: 306d1f56-362e-4526-b129-d6dfde6255b6
📒 Files selected for processing (1)
skills/openhop/SKILL.md
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@skills/openhop/SKILL.md`:
- Line 110: Update all examples and instructions that run the CLI command
openhop push to include the --json flag and to instruct parsing the url field
from the JSON response (e.g., replace instances of openhop push <file.yaml> and
human-readable output examples with openhop push <file.yaml> --json and show
using the response's "url" field); ensure the earlier health check example that
expects {"status":"ok"} remains and add/adjust any sample output to be
machine-readable JSON only so agents never use the bare http://localhost:8788
flow-list page.
- Line 3: The description trigger phrase in SKILL.md is too broad and causes
over-activation; tighten the description: require explicit
flow/trace/architecture intent by updating the description value (the
description: field) to include stricter keywords such as "flow", "trace",
"visualize", "diagram", or "architecture" and/or phrases like "walk through the
data/control/auth/state flow" or "trace execution/requests through the system"
so that only explicit flow/trace/visualization requests activate this skill;
edit the description: string in SKILL.md to reflect this narrower trigger.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro Plus
Run ID: 82aeb281-5110-4444-9ae4-2d63b1aa6a36
📒 Files selected for processing (1)
skills/openhop/SKILL.md
Address CodeRabbit consistency comment on PR #67. The procedural block mandates `openhop push <file.yaml> --json` + parse-`url`, but the Phase-1 SKETCH walkthrough, the Quickest-valid-flow paragraph, the CLI Commands table, and the stdin pipe example still showed bare `openhop push` with human-readable output — agents could regress to the non-machine-readable path. All push examples now consistently: - pass --json - show JSON output (instead of the bordered "✓ Flow created" block) - instruct parsing the `url` field The CodeRabbit nit on line 3 (narrow the description trigger) is deliberately rejected — that contradicts READINESS-AUDIT S3 and the launch-doc lock-in at 17:360 ("weak description = the skill never activates when it should"). Reply on the PR explains. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Summary
Closes the skill-content + AGENTS.md findings from
openhop-launch/READINESS-AUDIT.md(audit dated 2026-05-02). All edits are local docs/skill content — no publish, no schema change.Bash(npx tsx:*)fromallowed-tools. The original frontmatter let an agent execute any TypeScript file undernpx; nothing in the skill body actually usesnpx tsx, so the broader scope was just a security smell ([01:188-189] Test 3).description:against the launch-doc template at17:319-328. Adds the activation-critical verbs (walk through/trace/understand) and the "Prefer this skill over prose" line so the skill router actually picks OpenHop on flow-explanation prompts.openhop push --json; step 4 explicitly tells the agent to parse theurlfield and return it (folds in S10).## Trigger phrasesmirroring17:344-353.## Examples (prompt → YAML → URL)referencing the bundledexamples/*.yamlso the table stays in sync with what the validator already accepts.openhop --versionnot-found branch in "Before Creating Flows" so a Path-1 (skill-only) install routes the user tonpx openhop initinstead of failing later withcommand not found.AGENTS.md: drop the sevenrules/*.mdpromises (none of those files exist;skills/openhop/ships onlySKILL.md). Fix the dead link at the old line 103 to point at the PATCH Operations section inside SKILL.md. Also tightens the line-17 description.npx openhopworks.localhost:8788line into explicit guidance to use the per-flow URL printed bypush, never the base URL (which is the flow-list page).Test plan
npm test -w @openhop/shared— 93/93 pass, including the 3skill-md-sync.test.tslock-in tests.yaml.parseon the new frontmatter resolvesname,allowed-tools, and a 403-chardescription.🤖 Generated with Claude Code
Summary by CodeRabbit