.claude/skills/claude-code-action/SKILL.md
Knowledge base for creating and configuring Claude Code Action GitHub workflows
npx skillsauth add codyswanngt/lisa claude-code-actionInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Reference guide for creating anthropics/claude-code-action@v1 GitHub workflows.
Choose one authentication method:
| Method | Input | Use Case |
|--------|-------|----------|
| OAuth Token | claude_code_oauth_token | Recommended for most setups (requires Claude Pro or Max) |
| API Key | anthropic_api_key | Direct Anthropic API key from console.anthropic.com |
| AWS Bedrock | aws_access_key_id + aws_secret_access_key | AWS-hosted Claude |
| GCP Vertex | gcp_project_id + gcp_region + gcp_workload_identity_provider | Google Cloud Claude |
CLAUDE_CODE_OAUTH_TOKENRequires a Claude Pro or Max subscription.
claude setup-token
gh secret set CLAUDE_CODE_OAUTH_TOKEN
Paste the token when prompted.On macOS, Claude Code stores credentials in the encrypted Keychain (not a plain file). The setup-token command is the official way to extract a token for CI use.
| Name | Type | Required For | How to Set |
|------|------|-------------|------------|
| CLAUDE_CODE_OAUTH_TOKEN | Secret | All Claude workflows | gh secret set CLAUDE_CODE_OAUTH_TOKEN |
| ENABLE_CLAUDE_NIGHTLY | Variable | Nightly workflows (opt-in) | gh variable set ENABLE_CLAUDE_NIGHTLY --body "true" |
Triggered when users mention @claude in comments, reviews, or issues.
on:
issue_comment:
types: [created]
pull_request_review_comment:
types: [created]
issues:
types: [opened, assigned]
pull_request_review:
types: [submitted]
Triggered when a CI workflow fails. Automatically fixes the code.
on:
workflow_run:
workflows: ["CI Quality Checks"]
types: [completed]
Guard against infinite loops:
if: |
github.event.workflow_run.conclusion == 'failure' &&
!startsWith(github.event.workflow_run.head_branch, 'claude-auto-fix-') &&
github.event.workflow_run.head_branch != 'main' &&
github.event.workflow_run.head_branch != 'staging' &&
github.event.workflow_run.head_branch != 'dev'
Runs on a cron schedule for maintenance tasks (test improvement, coverage).
on:
schedule:
- cron: '0 3 * * 1-5' # 3 AM UTC weekdays
workflow_dispatch:
Use opt-in guard:
if: vars.ENABLE_CLAUDE_NIGHTLY == 'true'
permissions:
contents: write
pull-requests: write
issues: write
actions: read
id-token: write
Standard allowedTools for Lisa projects:
Edit,MultiEdit,Write,Read,Glob,Grep,Bash(git:*),Bash(npm:*),Bash(npx:*),Bash(bun:*),Bash(yarn:*),Bash(pnpm:*),Bash(gh:*)
This covers:
Bash(git:*) -- commit, push, branch, etc.Bash(gh:*) -- create PRs, issues, etc.| Input | Required | Description |
|-------|----------|-------------|
| prompt | No | Task instructions for Claude |
| claude_code_oauth_token | Yes* | OAuth token for authentication |
| claude_args | No | CLI args: --allowedTools, --max-turns, --system-prompt, --mcp-config |
| branch_prefix | No | Prefix for auto-created branches (e.g., claude/nightly-) |
| additional_permissions | No | Extra GitHub permissions (e.g., actions: read) |
| max_turns | No | Max agentic turns (via claude_args --max-turns) |
| track_progress | No | Enable progress tracking comments |
| allowed_bots | No | Comma-separated bot names allowed to trigger |
| allowed_non_write_users | No | Users without write access who can trigger |
Pass MCP server config via claude_args:
claude_args: |
--mcp-config .mcp.json
Pass secrets to MCP servers via environment variables in the workflow.
Before running nightly workflows, check for existing open PRs:
- name: Check for existing PR
id: check-pr
uses: actions/github-script@v7
with:
script: |
const pulls = await github.rest.pulls.list({
owner: context.repo.owner,
repo: context.repo.repo,
state: 'open',
per_page: 100,
});
const existing = pulls.data.find(pr =>
pr.head.ref.startsWith('claude/nightly-') &&
pr.title.toLowerCase().includes('your-keyword')
);
core.setOutput('has_existing_pr', existing ? 'true' : 'false');
- name: Run Claude
if: steps.check-pr.outputs.has_existing_pr != 'true'
uses: anthropics/claude-code-action@v1
Use --max-turns to limit API usage:
claude_args: |
--max-turns 25
Recommended limits:
${{ secrets.* }} for all sensitive valuesallowed_bots to control which bots can trigger Claudedevelopment
Prepare a machine — a fresh laptop or a throwaway container — to run coding agents, before any repository exists. Detects which of Lisa's supported agents (Claude Code, Codex, Cursor, OpenCode, Antigravity, Copilot) are already installed, asks which credential manager the machine uses (Bitwarden, 1Password, Doppler, Vault, AWS, or none), and installs only what is missing, each by its vendor's own preferred method. Idempotent, headless by default, and emits a Dockerfile for a spin-up/spin-down environment. Run it on a new machine, in a container, or before cloning anything.
tools
Provision and verify a remote execution environment for a host project — Codex Cloud today, other remote surfaces as they are added. Generates a repository-owned setup script that installs the declared toolchain, materializes secrets through lisa-secrets-access, and runs the project's own hook. Provisions by API where one exists, by driving the vendor console where one does not, and by emitting exact config otherwise — then proves the result with the same read-back regardless of which tier did the work. Use before dispatching any work with executionEnv.
tools
Bring a developer's machine in line with the toolchain the project declares. Reports every tool in remoteEnv.tools that is missing, outdated, or unpinned for this platform, and installs the missing ones into ~/.local/bin from the same pinned, checksummed entries the remote surfaces use — but only when asked. Same manifest, same pins, same installers as lisa-setup-remote-env; what differs is consent and that the pin is a floor rather than an equality. Run it on a fresh checkout, after a manifest change, or when a tool fails at the moment of use.
tools
Route one unit of work to a remote execution surface. Reads the executionEnv parameter (local by default, codex-cloud or claude-web today), verifies the environment is provisioned and bound to this repository, submits a thin skill invocation, records the task identifier to .lisa/remote-dispatch.json, and exits without polling. Routing only — the remote runs the identical skill from the identical repository. Composable and inline: other skills invoke it via the Skill tool rather than users calling it directly.