# Attensira

> Track whether AI assistants mention and cite your brand, and act on what you find.

- [Attensira documentation](https://docs.attensira.com/index.md): Track whether AI assistants mention and cite your brand, and put an agent to work on what you find.
- [What Attensira does](https://docs.attensira.com/start/what-it-does.md): The daily loop: your prompts go to real models, and we record who got mentioned and who got cited.
- [The free audit](https://docs.attensira.com/start/free-audit.md): What attensira.com/audit reads, what it shows everyone, and what one email address unlocks.
- [Quickstart](https://docs.attensira.com/start/quickstart.md): Add prompts, confirm competitors, and read your first measured number within a day.
- [Your first fifteen minutes](https://docs.attensira.com/start/first-fifteen-minutes.md): Choose your questions, pay, and watch the first readings land, then get the first thing to do.
- [Your first reading](https://docs.attensira.com/start/first-reading.md): What happens the moment your subscription goes active: a dedicated pass reads every tracked prompt right away, and files what it finds in your Inbox.
- [The set-up run](https://docs.attensira.com/start/set-up-run.md): The one session that runs itself once your first readings are in: it reads your site, groups what is wrong by the source that decided it, and files a starting plan.
- [Concepts and entity model](https://docs.attensira.com/start/concepts.md): Every word Attensira uses, what it means, and the wire field an MCP user actually sees.
- [Prompts](https://docs.attensira.com/tracking/prompts.md): Add the questions people actually ask assistants, and manage the prompt slots your plan gives you.
- [Topics](https://docs.attensira.com/tracking/topics.md): Group related prompts so reporting rolls up by theme instead of one question at a time.
- [Competitors](https://docs.attensira.com/tracking/competitors.md): Track rival domains alongside your own, and read competitor rates correctly — they do not sum to 100%.
- [Models](https://docs.attensira.com/tracking/models.md): Choose which assistant surfaces answer your prompts, and understand tracked versus readable.
- [How we collect data](https://docs.attensira.com/measure/how-we-collect-data.md): The full measurement methodology: which model surfaces are read, how often, how deeply we sample, and the significance test every delta has to pass.
- [Share of voice](https://docs.attensira.com/measure/share-of-voice.md): How often a model names your brand in its answer, measured as successful runs naming you divided by all successful runs in the window.
- [Citation rate](https://docs.attensira.com/measure/citation-rate.md): How often a model attaches a link to your own domain, measured as successful runs citing you divided by all successful runs in the window.
- [Citations and sources](https://docs.attensira.com/measure/citations-and-sources.md): The difference between a mention, a citation and a source, plus how source share and average citation position are calculated.
- [Who decides an answer](https://docs.attensira.com/measure/deciding-sources.md): For every prompt, on every model, the pages the answers actually leaned on — whether you are on each one, and the next move.
- [How to win one answer](https://docs.attensira.com/measure/win-this-answer.md): The win plan for a single prompt: per model, what decided the answer, which rung you are on, the one next action, and when we read it again.
- [Reading the numbers](https://docs.attensira.com/measure/reading-the-numbers.md): How to tell a measured zero from an unmeasured cell, why a rate carries a sample count, and when a change is real.
- [Pages](https://docs.attensira.com/traffic/pages.md): See which of your own URLs models attach as citations, how often, and where they land in the citation list.
- [AI traffic](https://docs.attensira.com/traffic/ai-traffic.md): AI crawlers fetch your pages; models answer prompts. Two systems, two numbers, and why measurement lives in edge middleware.
- [Chats](https://docs.attensira.com/agent/chats.md): Ask the agent a question about your workspace, see what data it can reach, and choose which agent model answers.
- [Tables, charts and answers](https://docs.attensira.com/agent/what-the-agent-shows.md): What the agent draws alongside its reply, why an unmeasured day is a gap and not a zero, and how to copy or download what it shows you.
- [Searching what was actually said](https://docs.attensira.com/agent/recall.md): Recall searches the answers we have already read for the passages nearest your question, and hands them back with the prompt, platform and day they came from.
- [Deep reads](https://docs.attensira.com/agent/deep-read.md): Four read-only tools that return the stored answers, citation days and page versions your metrics were counted from.
- [Automations](https://docs.attensira.com/agent/automations.md): A standing instruction the agent runs on a cadence, with a run history — and an idempotency key so a retry never bills twice.
- [What the agent notices on its own](https://docs.attensira.com/agent/findings.md): The hourly rules that watch your readings and file what changed — without a person and without a model asking a question.
- [Approval modes](https://docs.attensira.com/agent/approval.md): One workspace setting decides how much the agent does on its own before anything leaves the workspace — Ask, Draft or Act.
- [Inbox](https://docs.attensira.com/agent/inbox.md): Everything that needs you in one list: approvals the agent is waiting on, work it could not finish alone, and what it is doing right now.
- [The morning brief](https://docs.attensira.com/agent/morning-brief.md): One email a day, at 07:00 in your own time zone: what got done, what needs you, and what changed — with no comparison until there is enough data to make one honestly.
- [GitHub](https://docs.attensira.com/agent/github.md): Install the Attensira GitHub App and the agent opens pull requests against your site's repositories — it never merges them, and never pushes twice.
- [Off-site work](https://docs.attensira.com/agent/off-site.md): Most of what decides an AI answer is not on your site. Here is what the agent does about the pages, lists and repositories that are — pitches, entries in other people's repositories, and original pages worth citing.
- [Webflow and WordPress](https://docs.attensira.com/agent/cms.md): How Attensira changes a page when your site is not in a repository: a draft in your own editor, never a silent publish.
- [Knowledge and skills](https://docs.attensira.com/agent/knowledge-and-skills.md): Give the agent a brand profile and reusable skills so its output sounds like you and follows your rules.
- [What the agent remembers](https://docs.attensira.com/agent/what-we-know.md): The evidence-cited facts the agent derives about your workspace every night, so a conversation doesn't start by rediscovering what last night's run already knew.
- [Plugins](https://docs.attensira.com/agent/plugins.md): Connect third-party tools to the agent, or register your own MCP server so the agent can call it.
- [Attensira MCP server](https://docs.attensira.com/mcp/overview.md): Query your Attensira workspace and act on it from inside your AI assistant, over Streamable HTTP with OAuth or a workspace-scoped API key.
- [Connect a client](https://docs.attensira.com/mcp/connect.md): Complete, copy-pasteable Attensira MCP setup for Claude Code, Claude Desktop, Claude.ai, Cursor, VS Code, ChatGPT, Codex CLI, Windsurf, Zed, Cline, n8n, and any other MCP client.
- [OAuth for the MCP server](https://docs.attensira.com/mcp/oauth.md): Discovery, scopes, PKCE and token lifetimes for connecting a client to the Attensira MCP server.
- [MCP tool reference](https://docs.attensira.com/mcp/tools.md): All 33 Attensira MCP tools — parameters, return shapes, which credential each needs, and which ones spend credits or delete things.
- [MCP recipes](https://docs.attensira.com/mcp/recipes.md): Six task-shaped recipes for the Attensira MCP server — the call to make, the chain behind it, how to read the result, and what it cannot tell you.
- [MCP troubleshooting](https://docs.attensira.com/mcp/troubleshooting.md): Fix sign-ins that never return, 401s, scope errors, missing tools, revoked credentials, empty model results, and timeouts on the Attensira MCP server.
- [Attensira REST API](https://docs.attensira.com/api/overview.md): Read your workspace and change it from your own code, over conventional HTTP, using the same workspace-scoped API key as the MCP server.
- [Authenticating REST API requests](https://docs.attensira.com/api/authentication.md): Send a workspace-scoped API key as a bearer token. The key names the workspace, so no request can be pointed at someone else's.
- [REST API endpoint reference](https://docs.attensira.com/api/endpoints.md): All sixteen Attensira REST endpoints — parameters, request bodies, response shapes, and which ones spend credits or delete data.
- [REST API errors](https://docs.attensira.com/api/errors.md): Every failure is one envelope with a stable type, so a client needs exactly one branch for things going wrong.
- [Plans](https://docs.attensira.com/account/plans.md): What each Attensira plan includes, and how prompt slots and the monthly response budget differ.
- [Credits](https://docs.attensira.com/account/credits.md): What a credit is, which agent actions spend one, and what happens as a workspace balance falls to zero.
- [Your receipt](https://docs.attensira.com/account/receipt.md): One page that says what we actually did, what is waiting or failed, and whether we can prove anything moved — built to forward, not to interpret.
- [API keys](https://docs.attensira.com/account/api-keys.md): Mint an Attensira API key, choose its scope, store the one-time reveal, and revoke it when you are done.
- [Connected applications](https://docs.attensira.com/account/connections.md): How signing an AI client in to Attensira works, what it can reach, and how to disconnect it.
- [Email notifications](https://docs.attensira.com/account/notifications.md): What Attensira emails you about, and how to turn any category off without turning off the rest.
- [Limitations](https://docs.attensira.com/reference/limitations.md): What Attensira does not measure, where its methods are approximate, and how to read numbers accordingly.
- [Troubleshooting](https://docs.attensira.com/reference/troubleshooting.md): Answers to the questions people ask when an Attensira number looks wrong or missing.