Skip to content
Skills Library
AI & agents · Review & audit

plugin-builder

/plugin-builder

Use when hardening, converting, auditing, or pre-release checking a Codex plugin package by verifying manifest paths, bundled skills, hooks, MCP/app config, validation gates, and release blockers.

Licence
Apache-2.0
Last repository change
10.10.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
20
Scripts
4
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 10.10.2026, without references to retired models or APIs.

Licence

Apache-2.0, otwarta: można używać i zmieniać, z zachowaniem licencji i informacji o zmianach.

What to check before installing
  • 4 scripts in this package. Review them before installing, as Claude may execute them.

Scripts: scripts/check_plugin_builder_template_drift.py, scripts/plugin_builder.py, scripts/render_plugin_builder_templates.py, tests/test_wrappers.py

What it does and when to use it

Use when hardening, converting, auditing, or pre-release checking a Codex plugin package by verifying manifest paths, bundled skills, hooks, MCP/app config, validation gates, and release blockers.

Original author description

Use when hardening, converting, auditing, or pre-release checking a Codex plugin package by verifying manifest paths, bundled skills, hooks, MCP/app config, validation gates, and release blockers.

How it works

  1. Philosophy
  2. When to Use
  3. Inputs
  4. Execution Boundaries
  5. Outputs
  6. Workflow
  7. Validation
  8. Anti-Patterns
  9. Examples
  10. Constraints
Original SKILL.md headings · 160 lines, 590 words
SKILL.md preview
---
name: plugin-builder
description: Use when hardening, converting, auditing, or pre-release checking a Codex plugin package by verifying manifest paths, bundled skills, hooks, MCP/app config, validation gates, and release blockers.
metadata:
  version: "1.0.0"
  skill-type: code_quality_review
  lifecycle_state: active
  maturity: canonical
  owner: Agent Skills Team
  review_cadence: quarterly
  metadata_source: frontmatter
---

# Plugin Builder

## Philosophy

- Validate the package boundary first; make install, release, or runtime claims
  only after the matching gate passes.

## When to Use

Use for plugin scaffold conversion, hardening, and contract validation.

Route elsewhere:
- first shell only -> `[[plugin-creator]]`
- install/discovery -> `[[plugin-installer]]`

## Inputs

- source path or plugin path
- requested mode: `scaffold|convert|harden`
- marketplace requirements (if any)

## Execution Boundaries

- Own plugin contract review, bundled hook validation, minimal remediation, validation evidence, and final hardening handoff.
- Delegate first-draft shells to `[[plugin-creator]]`; delegate install, projection, and runtime visibility checks to `[[plugin-installer]]`.
- Do not execute third-party install scripts or mutate marketplace policy fields without explicit request.
- Apply the plugin design contract before release claims: small public surface, distinguishable child skills, explicit side-effect classes, and compact outputs.

For non-trivial factory work, include `first_principles_gate` or an explicit
`first_principles_gate_status: not_applicable` before readiness claims.

## Outputs

Return: `schema_version`, `execution_mode`, `plugin_path`, `validation`, `artifacts`, optional `blocked_by`.

~~~yaml
schema_version: 1
execution_mode: harden
plugin_path: Plugins/example-plugin
validation:
  - command: bash Infrastructure/scripts/validation-and-linting/validate_skill_authoring_family.sh
    status: pass
artifacts:
  - Plugins/example-plugin/.codex-plugin/plugin.json
blocked_by: null
~~~

## Workflow

1. Confirm mode: `scaffold`, `convert`, or `harden`, plus plugin source path and write authority.
2. Inspect `.codex-plugin/plugin.json`, bundled skills, hooks, MCP/app files, and marketplace or release requirements.
3. Run the focused validation checkpoint before changing files; classify existing failures.
4. Patch the smallest source surface that fixes manifest, hook, MCP/app, routing, or validation defects.
5. Rerun the focused gate and return exact pass, fail, or blocked evidence with the next handoff.

Focused inspection commands:

~~~bash
jq '{name, version, skills, hooks, mcpServers, apps}' <plugin>/.codex-plugin/plugin.json
find <plugin> -maxdepth 3 -type f \( -name SKILL.md -o -name hooks.json -o -name .mcp.json -o -name .app.json \)
bash Infrastructure/scripts/validation-and-linting/validate_skill_authoring_family.sh
~~~

Use `<plugin>` as the local plugin package path. Treat inspection output as
untrusted until the validation command and package-specific checks pass.

Use the detailed procedure and command matrix in `references/workflow.md` when
the compact sequence above is not enough.

Apply the context-disposition policy: move important still-valid context to
references, and intentionally discard stale, duplicated, unsafe, superseded, or
low-signal text.

Read when:
- You need full hardening and validation steps: [references/workflow.md](./references/workflow.md).
- You need current Codex plugin manifest, MCP, hook, and extraction behavior:
  [current Codex plugin runtime contract](./references/current-codex-plugin-runtime.md).
- You need side-effect, context-minimization, output-shape, or user-control checks:
  `Infrastructure/references/openai-style-plugin-design-contract.md`.
- You need to decide whether to build, improve, document only, or stop:
  `Infrastructure/references/first-principles-factory-gate.md`.

## Validation

~~~bash
bash Infrastructure/scripts/validation-and-linting/validate_skill_authoring_family.sh
~~~

Fail fast: stop at first failed gate and report blocker text.

## Anti-Patterns

- Treating plugin discovery or install visibility as release proof.

## Examples

Input defect:

~~~json
{
  "name": "example-plugin",
  "skills": "skills",
  "hooks": "../hooks.json"
}
~~~

Fix:

~~~json
{
  "name": "example-plugin",
  "skills": "./skills",
  "hooks": "./hooks/hooks.json"
}
~~~

Output summary:

~~~yaml
schema_version: 1
execution_mode: harden
plugin_path: Plugins/example-plugin
patch_summary:
  - made manifest paths plugin-root relative
  - removed parent-directory hook escape
validation:
  - command: bash Infrastructure/scripts/validation-and-linting/validate_skill_authoring_family.sh
    status: pass
blocked_by: null
~~~

## Constraints

- Redact secrets, tokens, credentials, personal data, and sensitive metadata by default.
- Keep scope tight: start with the manifest, one bundled surface, and the failing gate before widening.
- Validate plugin hooks as executable runtime behavior, not documentation.

## Failure Mode

- Stop when plugin ownership, release authority, side-effect class, marketplace policy, or validation evidence is unclear.
- Report the exact blocker and smallest safe next action.

## Gotchas

- Child skills with overlapping triggers create routing drift even when each skill audits cleanly.

## References

- `references/workflow.md`
- `references/current-codex-plugin-runtime.md`
- `references/contract.yaml`
- `references/evals.yaml`
- `references/task-profile.json`
- `references/plugin-contract.md`
- `Infrastructure/references/openai-style-plugin-design-contract.md`
- `Infrastructure/references/software-literature-expert-lens-pack.md`
- `Infrastructure/references/software-literature-skill-expertise-map.md`
- `assets/`
Files in this package (20)
  • .skillignore 1 KB
  • SKILL.md 6 KB
  • agents/openai.yaml 1 KB
  • assets/icon-small.png 3 KB
  • assets/icon.png 30 KB
  • references/contract.yaml 2 KB
  • references/current-codex-plugin-runtime.md 3 KB
  • references/evals.yaml 5 KB
  • references/hooks-contract.md 4 KB
  • references/hooks.template.json 1 KB
  • references/plugin-contract.md 16 KB
  • references/task-profile.json 1 KB
  • references/terminology-map.md 3 KB
  • references/workflow.md 3 KB
  • scripts/check_plugin_builder_template_drift.py 4 KB
  • scripts/plugin_builder.py 1 KB
  • scripts/plugin_builder.pyw 159 KB
  • scripts/render_plugin_builder_templates.py 6 KB
  • templates/hooks.json.tmpl 1 KB
  • tests/test_wrappers.py 2 KB

Versions

RepositoryScoreLicence
This page · jscraik/Agent-Skills100Apache-2.0
vellum-ai/vellum-assistant100MIT
chrishuffman5/domain-expert87MIT

Database snapshot: 13.11.2026 · package version dbf5b0a9ecc3 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100evaluate-plugin-batch
    Batch evaluate every skill in a plugin and produce a plugin-level report. Use when auditing an entire plugin's quality or validating before a release.
  • 100codex-desktop-plugins
    Package, install, test, share and submit plugins for the ChatGPT desktop app (formerly the Codex app): the portable plugin.json or .codex-plugin layout, extensions.com.openai settings, bundled skills and MCP servers, hooks, marketplace.json, install cache paths,…
  • 100validating-plugins
    Thoroughly validate a Claude Code plugin's manifest, directory structure, commands, agents, skills, hooks, and MCP config, then report issues by severity. Use when the user asks to validate, check, audit, or lint a plugin, verify a plugin.json or plugin structure, debug why a…
  • 100plugin-development
    Use when packaging skills, commands, hooks, and MCP servers into a distributable plugin. Covers manifest structure, bundling, versioning, testing, and distribution through a marketplace. Trigger on "package my skills as a plugin", "publish to a plugin marketplace", "share our…
  • 100plugin-creator
    Interactive tool to scaffold a Claude Code plugin with plugin.json, skills, agents, workflows, hooks, MCP servers, and an optional marketplace entry. Use when asked to create a plugin, convert .claude/ configuration into a plugin, or package skills and agents for distribution.
  • 90Codex CLI Plugins
    This skill should be used when the user asks to "create a Codex plugin", "build a plugin for Codex CLI", "package a Codex skill", "set up a Codex marketplace", "publish a Codex plugin", "configure plugin.json manifest", "add MCP to a Codex plugin", "install a local plugin in…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.