delimit

Lint two OpenAPI specs for breaking changes and policy violations. Primary CI integration point. Combines diff + policy into pass/fail.

Args: old_spec: Path to the old (baseline) OpenAPI spec file. new_spec: Path to the new (proposed) OpenAPI spec file. policy_file: Optional path to a .delimit/policies.yml file.

Diff two OpenAPI specs and list all changes. Pure diff, no policy.

Args: old_spec: Path to the old OpenAPI spec file. new_spec: Path to the new OpenAPI spec file.

Inspect or validate governance policy configuration.

Args: spec_files: List of spec file paths. policy_file: Optional custom policy file path.

Query the append-only contract ledger (hash-chained JSONL).

Args: ledger_path: Path to the events.jsonl ledger file. api_name: Filter events by API name. repository: Filter events by repository. validate_chain: Validate hash chain integrity.

Analyze downstream impact of an API change. Informational only.

Args: api_name: The API that changed. dependency_file: Optional path to dependency manifest.

Classify the semver bump for a spec change (MAJOR/MINOR/PATCH/NONE).

Deterministic classification based on diff engine output. Optionally computes the next version string.

Args: old_spec: Path to the old OpenAPI spec file. new_spec: Path to the new OpenAPI spec file. current_version: Optional current version (e.g. "1.2.3") to compute next version.

Generate a human-readable explanation of API changes.

7 templates: developer, team_lead, product, migration, changelog, pr_comment, slack.

Args: old_spec: Path to the old OpenAPI spec file. new_spec: Path to the new OpenAPI spec file. template: Template name (default: developer). old_version: Previous version string. new_version: New version string. api_name: API/service name for context.

Extract OpenAPI spec from framework source code (no spec file needed).

Detects the API framework (FastAPI, Express, NestJS) and extracts a complete OpenAPI specification directly from the source code. Currently supports FastAPI with full fidelity.

Args: project_dir: Path to the project root directory. python_bin: Optional Python binary path (auto-detected if omitted).

Initialize Delimit governance for a project. Creates .delimit/policies.yml and ledger directory.

Args: project_path: Project root directory. preset: Policy preset — strict, default, or relaxed.

Create a governed execution plan (Pro).

Args: operation: Operation to plan (e.g. "deploy", "migrate"). target: Target component or service. parameters: Operation parameters. require_approval: Whether to require approval before execution.

Get current Delimit OS status with plan/task/token counts (Pro).

Check governance gates for a plan (Pro).

Args: plan_id: The plan ID (e.g. "PLAN-A1B2C3D4").

Check governance system health.

Args: repo: Repository path to check.

Get current governance status for a repository.

Args: repo: Repository path.

Get governance policy for a repository (Pro).

Args: repo: Repository path.

Evaluate if governance is required for an action (requires governancegate) (Pro).

Args: action: The action to evaluate. context: Additional context. repo: Repository path.

Create a new governance task (requires governancegate) (Pro).

Args: title: Task title. scope: Task scope. risk_level: Risk level (low/medium/high/critical). repo: Repository path.

Run a governance task (requires governancegate) (Pro).

Args: task_id: Task ID to run. repo: Repository path.

Verify a governance task (requires governancegate) (Pro).

Args: task_id: Task ID to verify. repo: Repository path.

Search conversation memory semantically (Pro).

Args: query: Natural language search query. limit: Maximum results to return.

Store a memory entry for future retrieval (Pro).

Args: content: The content to remember. tags: Optional categorization tags. context: Optional context about when/why this was stored.

Get recent work summary from memory (Pro).

Args: limit: Number of recent entries to return.

Search vault entries (Pro).

Args: query: Search query for vault entries.

Check vault health status (Pro).

Get a vault state snapshot (Pro).

Plan deployment with build steps (Pro).

Args: app: Application name. env: Target environment (staging/production). git_ref: Git reference (branch, tag, or SHA).

Build Docker images with SHA tags (Pro).

Args: app: Application name. git_ref: Git reference.

Publish images to registry (Pro).

Args: app: Application name. git_ref: Git reference.

Rollback to previous SHA (Pro).

Args: app: Application name. env: Target environment. to_sha: SHA to rollback to.

Get deployment status (Pro).

Args: app: Application name. env: Target environment.

Register a new dataset in the file-based intel registry.

Args: name: Dataset name. schema: Optional JSON schema for the dataset. description: Human-readable description.

List all registered datasets from the intel registry.

Mark a dataset as immutable (frozen). Prevents further modifications.

Args: dataset_id: Dataset identifier.

Store a research snapshot with provenance metadata in the local intel store.

Args: data: Snapshot data (any JSON-serializable dict). provenance: Optional provenance metadata (source, author, etc.).

Search saved intel snapshots by keyword, date, or dataset.

Args: dataset_id: Optional dataset to filter by. query: Keyword search string. parameters: Optional params (date_from, date_to, limit).

Generate code template.

Args: template_type: Template type (component, page, api, etc.). name: Name for the generated code. framework: Target framework. features: Optional feature flags.

Scaffold new project structure.

Args: project_type: Project type (nextjs, api, library, etc.). name: Project name. packages: Packages to include.

Scan for security vulnerabilities.

Args: target: Repository or file path.

Audit security: dependency vulnerabilities, anti-patterns, and secret detection.

Scans for:

• Dependency vulnerabilities (pip-audit, npm audit)

• Hardcoded secrets (API keys, tokens, passwords)

• Dangerous patterns (eval, exec, SQL injection, XSS)

• .env files tracked in git

Optional: Set SNYK_TOKEN or install Trivy for enhanced scanning.

Args: target: Repository or file path to audit.

Collect evidence artifacts for governance (Pro).

Args: target: Repository or task path.

Verify evidence bundle integrity (Pro).

Args: bundle_id: Evidence bundle ID to verify. bundle_path: Path to evidence bundle file.

(Pro). Generate a release plan from git history.

Reads git log since last tag, counts commits and changed files, suggests a semver version, and generates a release checklist. Saves plan to ~/.delimit/deploys/ for tracking.

Args: environment: Target environment (staging/production). version: Release version (auto-detected if empty). repository: Repository path (default: current directory). services: Optional service list.

(Pro). Check release/deploy status from file-based tracker and git state.

Shows latest deploy plan, current git tag, how many commits HEAD is ahead of the tag, and recent deploy history.

Args: environment: Target environment (staging/production).

(Pro). Analyze project costs by scanning Dockerfiles, dependencies, and cloud configs.

Args: target: Project or infrastructure path to analyze.

(Pro). Find cost optimization opportunities: unused deps, oversized images, uncompressed assets.

Args: target: Project or infrastructure path to analyze.

(Pro). Manage cost alerts (file-based). CRUD operations on spending thresholds.

Args: action: Action (list/create/delete/toggle). name: Alert name (required for create). threshold: Cost threshold in USD (required for create). alert_id: Alert ID (required for delete/toggle).

Validate data files: JSON parse, CSV structure, SQLite integrity check.

Args: target: Directory or file path containing data files.

Check for migration files (alembic, Django, Prisma, Knex) and report status.

Args: target: Project path to scan for migration files.

Back up SQLite and JSON data files to ~/.delimit/backups/ with timestamp.

Args: target: Directory or file to back up.

(Pro). Query live system metrics (CPU, memory, disk I/O, network).

Query types: cpu, memory, disk, io, network, system (default), all. Reads directly from /proc for real-time data.

Optional: Set PROMETHEUS_URL for remote metrics.

Args: query: Metrics query type (cpu|memory|disk|io|network|system|all). time_range: Time range (e.g. "1h", "24h", "7d"). source: Optional metrics source (prometheus, local).

(Pro). Search system and application logs.

Searches journalctl, /var/log/*, and application log directories. Returns matching log lines with source attribution.

Optional: Set ELASTICSEARCH_URL or LOKI_URL for centralized log search.

Args: query: Log search query string. time_range: Time range (5m, 15m, 1h, 6h, 24h, 7d). source: Log source path or integration name (journalctl, elasticsearch).

(Pro). System health check: disk space, memory, running services, uptime.

Checks disk usage, memory, process count, load average, and probes common service ports (Node, PostgreSQL, Redis, Nginx, etc.). No external integration needed.

Extract design tokens from project CSS/SCSS/Tailwind config (or Figma if FIGMA_TOKEN set).

Args: figma_file_key: Optional Figma file key (uses Figma API if FIGMA_TOKEN env var is set). token_types: Token types to extract (colors, typography, spacing, breakpoints). project_path: Project directory to scan. Defaults to cwd.

Generate a React/Next.js component skeleton with props interface and Tailwind support.

Args: component_name: Component name (PascalCase). figma_node_id: Optional Figma node ID (reserved for future use). output_path: Output file path. Defaults to components//.tsx. project_path: Project root for Tailwind detection.

Read existing tailwind.config or generate one from detected CSS tokens.

Args: figma_file_key: Optional Figma file key (reserved for future use). output_path: Output file path for generated config. project_path: Project root to scan for existing config or CSS tokens.

Validate responsive design patterns via static CSS analysis.

Scans for media queries, viewport meta, mobile-first patterns, fixed widths.

Args: project_path: Project path to validate. check_types: Check types (breakpoints, containers, fluid-type, etc.).

Scan for React/Vue/Svelte components and generate a component catalog.

Args: project_path: Project path to scan. output_format: Output format (json/markdown).

Generate a .stories.tsx file for a component (no Storybook install required).

Args: component_path: Path to the component file. story_name: Custom story name. Defaults to component name. variants: Variants to generate. Defaults to [Default, WithChildren].

Run visual regression test -- screenshot with Playwright and compare to baseline.

Falls back to guidance if Playwright is not installed.

Args: url: URL to screenshot. project_path: Project path for baseline storage. threshold: Diff threshold (0.0-1.0).

Run WCAG accessibility checks by scanning HTML/JSX/TSX for common issues.

Checks: missing alt, missing labels, empty buttons, heading order, aria-hidden on focusable.

Args: project_path: Project path to scan. standards: Accessibility standard (WCAG2A/WCAG2AA/WCAG2AAA).

Generate test skeletons for source code.

Scans source files using AST parsing (Python) or regex (JS/TS), extracts public function signatures, and generates test file skeletons.

Args: project_path: Project path. source_files: Specific files to generate tests for. framework: Test framework (jest/pytest/vitest).

Run smoke tests for a project.

Detects the test framework (pytest/jest/vitest/mocha) from project config, runs the test suite, and parses pass/fail/error counts.

Args: project_path: Project path. test_suite: Specific test suite to run.

Generate API reference documentation for a project.

Scans Python files for docstrings and JS/TS files for JSDoc comments. Produces a markdown API reference organized by source file.

Args: target: Project path.

Validate documentation quality and completeness.

Checks README existence, docstring coverage on public functions, and broken internal links in markdown files.

Args: target: Project path.

Check a GitHub issue for new comments since the last check.

Sensor tool for monitoring outreach issues. Returns a structured signal with new comments, issue state, and severity classification.

Args: repo: GitHub repository in owner/repo format (e.g. "activepieces/activepieces"). issue_number: The issue number to monitor. since_comment_id: Last seen comment ID. Pass 0 to get all comments.

Return Delimit unified server version, all tiers, and tool count.

Get help for a Delimit tool — what it does, parameters, and examples.

Args: tool_name: Tool name (e.g. 'lint', 'gov_health'). Leave empty for overview.

Diagnose your Delimit setup — check environment, config, and tool status.

Universal 'get me unstuck' command. Checks Python, MCP config, governance state, and reports any issues with suggested fixes.

Args: project_path: Project to diagnose.

Activate a Delimit Pro license key.

Args: license_key: The license key to activate (e.g. DELIMIT-XXXX-XXXX-XXXX).

Check current Delimit license status -- tier, validity, and expiry.

(Pro). Deploy a site — git commit, push, Vercel build, and deploy in one step.

Handles the full chain: stages changes, commits, pushes to remote, builds with Vercel, deploys to production. No manual steps needed.

Args: project_path: Path to the site project (must have .vercel/ configured). message: Git commit message. Auto-generated if empty.

(Pro). Publish an npm package — bump version, publish to registry, verify.

Full chain: check auth, bump version, npm publish, verify on registry, git commit + push the version bump. Use dry_run=true to preview first.

Args: project_path: Path to the npm package (must have package.json). bump: Version bump type — "patch", "minor", or "major". tag: npm dist-tag (default "latest"). dry_run: If true, preview without actually publishing.

Add a new item to a project's ledger.

The ledger tracks what needs to be done across sessions. Specify the venture/project name or path. If empty, auto-detects from current directory.

Args: title: What needs to be done. venture: Project name or path (e.g. "my-project", "./path/to/project"). Auto-detects if empty. ledger: "ops" (tasks, bugs, features) or "strategy" (decisions, direction). item_type: task, fix, feat, strategy, consensus. priority: P0 (urgent), P1 (important), P2 (nice to have). description: Details. source: Where this came from (session, consensus, focus-group, etc).

Mark a ledger item as done.

Args: item_id: The item ID (e.g. LED-001 or STR-001). note: Optional completion note. venture: Project name or path. Auto-detects if empty.

List ledger items for a venture/project.

Args: venture: Project name or path. Auto-detects if empty. ledger: "ops", "strategy", or "both". status: Filter by status — "open", "done", "in_progress", or empty for all. priority: Filter by priority — "P0", "P1", "P2", or empty for all. limit: Max items to return.

Get a quick summary of what's open in the ledger.

Auto-detects the venture from context. Pass a venture name to check a specific project. Returns the top 5 open items by priority so the AI knows what to work on.

Args: venture: Project name or path. Auto-detects if empty.

List all registered ventures/projects that Delimit has been used with.

Ventures are auto-registered when you use any Delimit tool in a project directory.

(Pro). View and configure AI models for multi-model deliberation.

Actions:

• "list": show configured models and what's available

• "detect": auto-detect API keys from environment and configure

• "add": add a model provider (set provider + api_key)

• "remove": remove a model provider (set provider)

Supported providers: grok, gemini, openai, anthropic, codex

Args: action: list, detect, add, or remove. provider: Model provider for add/remove (grok, gemini, openai, anthropic, codex). api_key: API key for the provider (only used with action=add). model_name: Optional model name override (e.g. "gpt-4o", "claude-sonnet-4-5-20250514").

Run multi-model consensus via real AI-to-AI deliberation (Pro).

Models (Grok 4, Gemini, Codex) debate each other directly until unanimous agreement.

Args: question: The question to reach consensus on. context: Background context for all models. mode: "dialogue" (short turns) or "debate" (long essays). max_rounds: Maximum rounds (default 3 for debate, 6 for dialogue). save_path: Optional file path to save the full transcript.

(Pro). Audit or sync all public surfaces for consistency.

Checks GitHub repos, npm, site meta tags, CLAUDE.md, and releases against a central config. Reports what's stale and what needs updating.

Args: action: "audit" to check all surfaces, "config" to view/edit the release config.

Scan a project and show what Delimit can do for it.

First-run discovery tool. Finds OpenAPI specs, checks for security issues, detects frameworks, and suggests what to track. Use this when you first install Delimit or open a new project.

Args: project_path: Path to the project to scan.