AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Contribute

skill-hec-ovi-agentickit-contribute · by hec-ovi

|

No reviews yet
0 installs
19 views
0.0% view→install

Install

$ agentstack add skill-hec-ovi-agentickit-contribute

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets Used
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-hec-ovi-agentickit-contribute)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
2mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Contribute? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Contribute

Contract

By the end of this skill a contributor has:

  • A change localized to the right file under packages/agentickit/src/.
  • A matching test under the same package.
  • Updated public docs: both READMEs when the surface changes, plus the

relevant .pilot/skills/* if the procedure for consumers shifts.

  • A green pnpm --filter @hec-ovi/agentickit build and `pnpm --filter agentickit

test`.

  • A commit message matching conventions/commit-style.md.
  • Awareness of the rough size budget. The tree is ~5 kLoC today. A PR

that adds 500 net lines is a red flag worth justifying; a PR that shrinks is a celebration.

Iron Law: the public surface is the four hooks + two components + handler

The entire external API is:

  • usePilotState, usePilotAction, usePilotForm
  • `, , `
  • createPilotHandler, loadPilotProtocol, autoDetectModel (server)
  • parseResolver, parseSkill (protocol)
  • agentickit CLI (init, add-skill)

Adding to this surface is a deliberate choice with ecosystem consequences. Every new export is a promise you'll maintain it. Shrink before you grow; refactor before you add.

Phases

Phase 1: scope the change

Before editing, write down in one paragraph:

  • What the change enables that wasn't possible before.
  • Whether it extends the public surface (new export) or refines an

internal (same surface, different behavior).

  • What file(s) will change.
  • What test(s) will change.
  • What doc sections will change.

If that paragraph is vague, the change isn't ready to code.

Phase 2: find the right file

The layout under packages/agentickit/src/ is narrow on purpose:

| Concern | File | |---------|------| | Public exports | index.ts | | Client-side types | types.ts | | React contexts | context.ts | | Dev-mode helpers | env.ts | | A hook | hooks/use-pilot-*.ts | | The provider | components/pilot-provider.tsx | | The sidebar (split) | components/pilot-sidebar*.tsx | | The server handler | server/handler.ts | | Server entry point | server/index.ts | | .pilot/ parsing | protocol/resolver.ts, protocol/skill.ts, protocol/manifest.ts |

Prefer editing an existing file over creating a new one. Every new file is a new boundary someone has to reason about.

Phase 3: write the test first

Vitest + happy-dom. Tests live adjacent to source (or under a __tests__/ sibling if preferred). Run locally with pnpm --filter @hec-ovi/agentickit test:watch.

Test shape we want:

  • Happy path (canonical usage).
  • Edge case (empty input, cleanup on unmount, two registrations).
  • Failure mode (invalid schema, missing provider context).

Phase 4: implement

Match the existing style:

  • Opinionated TypeScript, no any.
  • Comments describe why, not what.
  • Zod at trust boundaries.
  • Refs for volatile values to avoid thrashing registrations (see

use-pilot-action.ts lines 57-64 for the canonical pattern).

  • Dev-mode warnings for misuse (see isDev() in env.ts and usage

throughout the hooks).

Phase 5: update docs when the surface moves

When you change the public surface, update in this order:

  1. README.md at the repo root (the narrative).
  2. packages/agentickit/README.md (the npm description, often shorter).
  3. .pilot/skills/*/SKILL.md for every skill whose procedure drifts.
  4. Type exports in packages/agentickit/src/index.ts (and

server/index.ts, protocol/index.ts).

Skip steps 1-2 for internal refactors that don't change the surface. Always do (3); the skills are the agent-facing contract.

Phase 6: validate

pnpm --filter @hec-ovi/agentickit build
pnpm --filter @hec-ovi/agentickit test
pnpm --filter @agentickit-examples/todo dev   # smoke test in the example

All three green. Then, and only then, stage and commit.

Phase 7: commit and push

Follow .pilot/conventions/commit-style.md. Quick template:

git add packages/agentickit/src/hooks/use-pilot-action.ts \
        packages/agentickit/src/hooks/__tests__/use-pilot-action.test.ts \
        README.md \
        packages/agentickit/README.md \
        .pilot/skills/register-action/SKILL.md

git commit -m "feat(hooks): "

Do NOT use git add -A. A stray .env, screenshot, or node_modules slipping in is a headache.

Phase 8: size check

Current package source (including CSS-in-JS + the CLI) sits at about 5 kLoC. After a change, eyeball the delta with:

cloc packages/agentickit/src   # or any line-counter

If a PR is adding a lot, ask: is there an existing file that shrinks? A comment that's too long? A helper that could be inlined? The "readable in an afternoon" promise is worth protecting.

Anti-Patterns

  • Adding a new peer dependency casually. Every peer is a prompt for

consumers to npm install something. The bar is high.

  • Introducing a wrapper around an AI SDK type that just re-exports it.

Consumers can import from ai directly.

  • Writing docs before the code exists. READMEs lie when they describe

vapor features. Ship the code first, then document it.

  • Fixing a bug in the example and not in the package. The example is

consumer code; the source of truth lives under packages/agentickit/src/.

  • Editing dist/. It's generated.

Output Format

After contributing, report:

  • The change summary (1 sentence).
  • The files touched (paths).
  • Test coverage added (which test file + how many cases).
  • Doc sections updated.
  • LoC delta for the package.
  • Proof that build and test are green.

Tools Used

  • pnpm --filter @hec-ovi/agentickit build / test / test:watch.
  • Edit source, test, and doc files.
  • git add , git commit -m.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.