# CLI commands URL: https://pagegraph.arya.sh/docs/reference/cli Every pagegraph CLI command with a one-line description, the agentic workflow command table, and the stdout and stderr convention. The `pagegraph` binary runs on [Bun](https://bun.sh) (`bunx pagegraph`). It bundles PageGraph's Effect, TypeSafe provider runtime and Pi agent runtime, so it does not depend on the app's Effect version. `vite` stays a peer, because graph commands load your app through Vite at runtime, and `lighthouse` is only needed by `pagegraph audit`. The graph commands acquire your graph through `pagegraph.config.ts` at the app root. Rename an existing `seo.config.ts`, `.js` or `.mjs` to its `pagegraph.config.*` equivalent. ## Graph commands [#graph-commands] ```bash pagegraph check # CI gate — exit 1 on structural violations pagegraph check --site # also check every declared page's rendered HTML pagegraph stale # refresh queue: sitemap pages past the freshness policy pagegraph graph # the graph as a tree · --format mermaid | json pagegraph inspect /pricing # one node: policy, sitemap status, in/out edges pagegraph inspect --live # fetch a deployed page, validate its rendered pagegraph links verify # crawl served HTML: depth, orphans, declared-vs-rendered pagegraph links verify --assert-coverage # also assert pagegraph.config.ts coverage on served anchors pagegraph links verify --emit-rendered # save the rendered edge set pagegraph links candidates # propose contextual links from the declared graph pagegraph research keywords --query "email api" # combine graph, Executor, and Jev evidence pagegraph improve metadata --page /pricing # apply a clean-tree source improvement pagegraph apply # apply a dry-run's recorded edits pagegraph sitemap # print sitemap.xml pagegraph robots # print robots.txt ``` | Command | What it does | | -------------------------------- | ---------------------------------------------------------------------------- | | `check` | Runs every rule on the declared graph; exits 1 on structural violations | | `check --site ` | Also checks the server-rendered HTML of every declared page | | `stale` | Lists sitemap-eligible pages past the freshness policy, oldest first | | `graph` | Prints the graph as a tree, or with `--format mermaid` or `json` | | `inspect ` | Shows one node: policy, sitemap status, incoming and outgoing edges | | `inspect --live` | Fetches a deployed page and validates its rendered `` | | `links verify ` | Crawls served HTML for depth, orphans and the declared-vs-rendered link diff | | `links candidates` | Proposes contextual links from the declared graph | | `sitemap` | Prints `sitemap.xml` | | `robots` | Prints `robots.txt` | | `apply ` | Applies the recorded edits of a dry run | | `diff ` | Compares two audit reports by semantic SEO outcome | | `audit ` | Audits any website without `pagegraph.config.ts` | | `init` | Scaffolds the shared agent preset under `.pagegraph/agent` | | `claims check` | Asks for missing claims answers and replays committed ones | | `markdown show`, `find`, `lock` | Inspects captured Markdown documents and maintains `markdown.lock.json` | | `skills list`, `skills get core` | Prints the bundled TanStack Start integration skill | ## Workflow commands [#workflow-commands] These run an agent through Pi. See [Run agent workflows](/docs/guides/agent-workflows) for setup. | Command | Result | | ---------------------- | -------------------------------------------------------------------- | | `research keywords` | Query demand, intent, and page ownership opportunities | | `research competitors` | Competitor topics, pages, and positioning evidence | | `research authority` | Qualified editorial, directory, partner, and community opportunities | | `analyze serp` | Search-result intent and page-shape fit | | `analyze content` | Substance, answer quality, and intent overlap | | `analyze ai-search` | Citation readiness and answer-engine coverage | | `plan architecture` | A page-backed information-architecture plan | | `improve content` | Evidence-backed source copy edits | | `improve metadata` | Title and description edits at the owning source | | `improve schema` | Structured-data edits justified by visible content | | `improve links` | Contextual internal-link edits from graph and page evidence | The four `improve` commands apply edits by default and require a clean Git tree. `--dry-run` records edits without changing files, and `--allow-dirty` applies them alongside existing changes. ## Agent skill [#agent-skill] ```bash pagegraph skills list pagegraph skills get core ``` The skill is maintained at [skill-data/core/SKILL.md](https://github.com/aryasaatvik/pagegraph/blob/main/skill-data/core/SKILL.md) and compiled into the CLI, so the instructions match the installed PageGraph version. `skills get core` prints Markdown to stdout, and `--json` returns a structured payload. ## Output convention [#output-convention] Stdout is data, stderr is status. `pagegraph check --json | jq` just works. Check the exit code as well as the output. ## Common flags [#common-flags] | Flag | Applies to | Meaning | | -------------------------------- | ------------------------------------------------- | ---------------------------------------------------------------------- | | `--json` | most commands | Emit versioned structured output on stdout | | `--site ` | `check`, `links candidates` | Fetch rendered pages from a running server | | `--allow-private` | `check --site`, `audit`, `links`, `improve links` | Permit private and reserved destinations; for local development and CI | | `--require-inbound "="` | `check` | Override the `coverage` policy for one run | | `--limit` | `links verify`, `links candidates`, workflows | Bound the crawl, plan or result count | | `--rendered ` | `links verify`, `links candidates` | Replay a saved rendered edge set instead of crawling | | `--from ` | workflow commands | Resume a prior run | | `--dev ` | `markdown`, `claims check` | Use the development bundle instead of the build | | `--refresh` | `claims check`, workflows | Ask again and replace cached answers |