0d3cb498a3
CI / Shell Format Check (push) Has been cancelled
CI / Check Ruby (3.4) (push) Has been cancelled
CI / CI Config (push) Has been cancelled
CI / Test on Node ${{ matrix.node }} and ${{ matrix.os }}${{ matrix.shard && format(' (shard {0}/3)', matrix.shard) || '' }} (push) Has been cancelled
CI / Build on Node ${{ matrix.node }} (push) Has been cancelled
CI / Style Check (push) Has been cancelled
CI / Generate Assets (push) Has been cancelled
CI / Check Python (3.14) (push) Has been cancelled
CI / Check Python (3.9) (push) Has been cancelled
CI / Build Docs (push) Has been cancelled
CI / Code Scan Action (push) Has been cancelled
CI / Site tests (push) Has been cancelled
CI / webui tests (push) Has been cancelled
CI / Run Integration Tests (push) Has been cancelled
CI / Run Smoke Tests (push) Has been cancelled
CI / Go Tests (push) Has been cancelled
CI / Share Test (push) Has been cancelled
CI / Redteam (Production API) (push) Has been cancelled
CI / Redteam (Staging API) (push) Has been cancelled
CI / GitHub Actions Lint (push) Has been cancelled
CI / Check Ruby (3.0) (push) Has been cancelled
release-please / release-please (push) Has been cancelled
release-please / build (push) Has been cancelled
release-please / publish-npm (push) Has been cancelled
release-please / publish-npm-backfill (push) Has been cancelled
release-please / docker (push) Has been cancelled
release-please / publish-code-scan-action (push) Has been cancelled
release-please / attest-code-scan-action (push) Has been cancelled
Deploy local.promptfoo.app / Deploy to Cloudflare Pages (push) Has been cancelled
Test and Publish Multi-arch Docker Image / test (push) Has been cancelled
Test and Publish Multi-arch Docker Image / build-docker-and-push-digests (map[digest-suffix:linux-amd64 platform:linux/amd64 runner:ubuntu-latest]) (push) Has been cancelled
Test and Publish Multi-arch Docker Image / build-docker-and-push-digests (map[digest-suffix:linux-arm64 platform:linux/arm64 runner:ubuntu-24.04-arm]) (push) Has been cancelled
Test and Publish Multi-arch Docker Image / merge-docker-digests (push) Has been cancelled
Test and Publish Multi-arch Docker Image / Attest Multi-arch Image (push) Has been cancelled
Validate Renovate Config / Validate Renovate Configuration (push) Has been cancelled
74 lines
3.1 KiB
Markdown
74 lines
3.1 KiB
Markdown
# Agent Plugin Bundles
|
|
|
|
This directory contains the Promptfoo agent plugin bundle. Read this file before
|
|
editing `plugins/promptfoo` or adding another plugin bundle.
|
|
|
|
## Promptfoo Plugin (shared Codex + Claude Code bundle)
|
|
|
|
`plugins/promptfoo` is a single bundle published to BOTH marketplaces:
|
|
|
|
- Codex, via `.codex-plugin/plugin.json` and `.agents/plugins/marketplace.json`.
|
|
- Claude Code, via `.claude-plugin/plugin.json` and the repo-root
|
|
`.claude-plugin/marketplace.json`.
|
|
|
|
One folder serves both because the manifest sidecar directories are disjoint and
|
|
both platforms auto-discover skills from `skills/`. The Codex-only
|
|
`agents/openai.yaml` files are ignored by Claude Code. Keep both manifests'
|
|
`name` fields equal (`promptfoo`) so the install identity matches across tools.
|
|
Keep both manifests' `version` fields equal, and bump them together whenever the
|
|
published bundle content changes. When a plugin declares an explicit version,
|
|
Claude Code only delivers changed plugin content after that version changes.
|
|
|
|
Keep the public surface to the four focused skills unless the product decision
|
|
changes:
|
|
|
|
- `promptfoo-evals`
|
|
- `promptfoo-provider-setup`
|
|
- `promptfoo-redteam-setup`
|
|
- `promptfoo-redteam-run`
|
|
|
|
Do not add a meta selector skill. Each platform routes from each skill's
|
|
frontmatter description (and, for Codex, the `agents/openai.yaml` default
|
|
prompt).
|
|
|
|
The standalone `plugins/promptfoo-evals` single-skill Claude bundle has been
|
|
retired; its eval skill is now the shared bundle's `promptfoo-evals`.
|
|
|
|
## Skill Layout
|
|
|
|
Each skill should keep:
|
|
|
|
- `SKILL.md` for concise workflow instructions and routing boundaries
|
|
- `agents/openai.yaml` for UI metadata
|
|
- `references/*.md` for concrete examples and longer patterns
|
|
- `scripts/*` only when deterministic helper code is useful
|
|
|
|
Keep examples one reference hop away from `SKILL.md`; avoid README-style files
|
|
inside skill folders.
|
|
|
|
## Provider And Redteam Assumptions
|
|
|
|
- Treat Python providers as first-class alongside JavaScript providers.
|
|
- Cover `file://provider.py` and `file://provider.py:function_name` wherever
|
|
provider examples mention local code.
|
|
- Redteam setup/run skills assume remote generation is available. Do not add
|
|
local-only or `PROMPTFOO_DISABLE_REMOTE_GENERATION` guidance unless the
|
|
product direction changes.
|
|
- Deterministic local providers are acceptable for fixtures and QA examples,
|
|
but the skills should steer real scans toward Promptfoo's normal generation
|
|
and grading paths.
|
|
|
|
## Validation
|
|
|
|
From the repo root:
|
|
|
|
```bash
|
|
CODEX_HOME="${CODEX_HOME:-$HOME/.codex}"
|
|
python3 "$CODEX_HOME/skills/.system/skill-creator/scripts/quick_validate.py" plugins/promptfoo/skills/promptfoo-evals
|
|
python3 "$CODEX_HOME/skills/.system/skill-creator/scripts/quick_validate.py" plugins/promptfoo/skills/promptfoo-provider-setup
|
|
python3 "$CODEX_HOME/skills/.system/skill-creator/scripts/quick_validate.py" plugins/promptfoo/skills/promptfoo-redteam-setup
|
|
python3 "$CODEX_HOME/skills/.system/skill-creator/scripts/quick_validate.py" plugins/promptfoo/skills/promptfoo-redteam-run
|
|
npx vitest test/agentSkills/promptfooPlugin.test.ts --run
|
|
npm run f
|
|
```
|