docs(skill): broaden triggers so OpenHop fires on diagram requests too#143
Merged
Conversation
The skill was framed too narrowly around code walkthroughs: an agent
that got "create a diagram explaining Orion features" wouldn't load
OpenHop, and would fall back to prose or static Mermaid.
This widens the description and the trigger phrase list to cover:
- explicit diagram requests ("create/draw/make a diagram of …")
- product / feature / architecture explainers, not just codebases
- user journeys, onboarding, integrations, pipelines, state machines
Adds a sniff test for the agent ("can the answer be expressed as a
sequence of named hops between named components?") so anything fitting
that shape routes to OpenHop, regardless of whether "code" was
mentioned. Also adds a product-flavored row to the prompt→YAML table.
WalkthroughThis PR broadens the OpenHop skill documentation from code-only walkthroughs to step-by-step flows for any system, product, or codebase, adds explicit constraints against prose/static diagrams for flow outputs, reorganizes trigger phrases (including diagram/visualization categories), and expands example patterns. ChangesOpenHop Skill Scope Expansion
Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 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)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@skills/openhop/SKILL.md`:
- Line 3: The description trigger in SKILL.md is too broad and will
over-activate OpenHop; narrow it by restricting the `description:` string to
explicit diagram/flow intents and/or require mention of named hops/components
(e.g., include phrases like "diagram", "draw", "architecture diagram", "flow",
"sequence of hops", or "visualize components and connections") instead of
generic explainer prompts like "how does V work"; update the `description:`
value in skills/openhop/SKILL.md (the description field) to only match explicit
diagram/flow requests or add an additional gating clause requiring named
components to avoid unintended activation.
🪄 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: b08c091c-c32e-49db-8eef-ccdecea35d55
📒 Files selected for processing (1)
skills/openhop/SKILL.md
Per review on PR #143: the previous description listed generic explainer phrasings like "how does V work" and "explain Z's features", which would route plain questions ("how does TCP work?", "explain closures") into OpenHop even though there's no flow to draw. Rewrites the description to: - Lead with the noun (render an animated diagram) instead of the generic verb (walk the user through). - Drop "how does V work" / "explain Z's features" trigger examples. - Add an explicit negative gate: do not activate for generic explainer questions with no named components or step sequence. - Keep the broad domain coverage (system / feature / codebase / workflow / pipeline / user journey) so explicit diagram intent still routes here.
Owner
Author
|
Addressed in
Body trigger lists are unchanged: once the skill activates, the body's examples help the agent shape the YAML — they don't gate activation. |
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@skills/openhop/SKILL.md`:
- Around line 44-46: The trigger phrases in SKILL.md (the description field
lines containing "how does {Product|Service|Platform} work?", "explain {Product
X}'s features", "walk me through how {users|customers} use {X}") are too generic
and cause over-activation; update those entries to require explicit
flow/step-by-step intent or named hop/diagram cues (e.g., prepend or append
tokens like "step-by-step", "flow", "diagram", or require a named hop component)
so the OpenHop skill only activates for visualization/stepwise requests;
specifically edit the description triggers in SKILL.md to replace the generic
forms with gated forms such as "how does {Product|Service|Platform} work —
step-by-step" or "walk me through how {users|customers} use {X}
(flow/diagram/hop:{name})" and remove or narrow any catch-all patterns.
🪄 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: 6c8a328e-7dc7-4db6-8f91-669a1c256a5e
📒 Files selected for processing (1)
skills/openhop/SKILL.md
5 tasks
naorsabag
added a commit
that referenced
this pull request
May 13, 2026
Cuts the first release since v0.2.0 on 2026-05-11. The publish workflow auto-publishes each package whose package.json version differs from the registry, then tags v0.3.0 and creates the GitHub release. Versions: - @openhop/server 0.2.0 -> 0.3.0 - @openhop/web 0.2.0 -> 0.3.0 - openhop (CLI) 0.2.0 -> 0.3.0 - CLI's pinned deps on server + web also bumped to 0.3.0. CHANGELOG: promotes the existing [Unreleased] block to [0.3.0] with the two user-visible shipping changes since v0.2.0: - Skill triggers broadened (#143) — diagram / visualization / walkthrough prompts about products, features, workflows, user journeys, etc. now route to OpenHop, not just code walkthroughs. Adds an explicit negative gate to keep generic explainer prompts ("how does TCP work?") from over-activating the skill. - INSPECT panel defaults to bottom-dock on mobile (#142) — narrow viewports keep the canvas full-width when the user opens the inspector. plumbing and not changelogged. #140 / #141 only touched the README and do not ship in any tarball.
naorsabag
added a commit
that referenced
this pull request
May 13, 2026
* chore: bump @openhop/server @openhop/web openhop to 0.3.0 Cuts the first release since v0.2.0 on 2026-05-11. The publish workflow auto-publishes each package whose package.json version differs from the registry, then tags v0.3.0 and creates the GitHub release. Versions: - @openhop/server 0.2.0 -> 0.3.0 - @openhop/web 0.2.0 -> 0.3.0 - openhop (CLI) 0.2.0 -> 0.3.0 - CLI's pinned deps on server + web also bumped to 0.3.0. CHANGELOG: promotes the existing [Unreleased] block to [0.3.0] with the two user-visible shipping changes since v0.2.0: - Skill triggers broadened (#143) — diagram / visualization / walkthrough prompts about products, features, workflows, user journeys, etc. now route to OpenHop, not just code walkthroughs. Adds an explicit negative gate to keep generic explainer prompts ("how does TCP work?") from over-activating the skill. - INSPECT panel defaults to bottom-dock on mobile (#142) — narrow viewports keep the canvas full-width when the user opens the inspector. plumbing and not changelogged. #140 / #141 only touched the README and do not ship in any tarball. * docs(changelog): add [0.3.0] link ref + point [Unreleased] at v0.3.0
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
1 participant
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
Real-world failure mode: a user asked their agent to "create a diagram that explains Orion features" and the agent fell back to prose / static Mermaid because the skill's trigger language was framed almost entirely around code walkthroughs.
This PR widens the surface so OpenHop loads on:
Specifically:
descriptionrewritten to drop the "flows through code" anchor and explicitly mention features, products, services, workflows, user journeys, integrations.No code changes; the validator, schema, and renderer are untouched. This is purely a routing/intent change for the agent.
Test plan
npx prettier --check skills/openhop/SKILL.mdclean🤖 Generated with Claude Code
Summary by CodeRabbit