{"query":"Agent instructions","corpusVersion":"local","generatedAt":"2026-09-13T04:40:10.634Z","results":[{"blockId":"agent-instructions#agent-instructions","pageId":"agent-instructions","title":"Agent instructions","pageTitle":"Agent instructions","url":"https://teammately.ai/docs/agent-instructions.md","humanUrl":"https://teammately.ai/docs/agent-instructions#agent-instructions","markdownUrl":"https://teammately.ai/docs/agent-instructions.md","sectionId":"agent-instructions","kind":"reference","productArea":"reference","score":891.5995514957898,"reasons":["search_match","title_match","term_match","prefix_or_fuzzy_match"],"markdown":"# Agent instructions\n\nUse this page as the behavioral policy for agents operating with Teammately docs. For route selection and context packaging, read [Agent context index](/docs/agent-context)."},{"blockId":"agent-instructions#related-reference-docs","pageId":"agent-instructions","title":"Related reference docs","pageTitle":"Agent instructions","url":"https://teammately.ai/docs/agent-instructions.md","humanUrl":"https://teammately.ai/docs/agent-instructions#related-reference-docs","markdownUrl":"https://teammately.ai/docs/agent-instructions.md","sectionId":"related-reference-docs","kind":"reference","productArea":"reference","score":361.2469638230087,"reasons":["search_match","page_title_match","term_match","prefix_or_fuzzy_match"],"markdown":"## Related reference docs\n\n{% related-card-grid title=\"Related reference docs\" %}\n- [Agent context index](/docs/agent-context)\n- [Reference index](/docs/reference)\n- [Object model](/docs/object-model)\n{% /related-card-grid %}"},{"blockId":"agent-instructions#required-context","pageId":"agent-instructions","title":"Required context","pageTitle":"Agent instructions","url":"https://teammately.ai/docs/agent-instructions.md","humanUrl":"https://teammately.ai/docs/agent-instructions#required-context","markdownUrl":"https://teammately.ai/docs/agent-instructions.md","sectionId":"required-context","kind":"reference","productArea":"reference","score":359.16519204263096,"reasons":["search_match","page_title_match","term_match","prefix_or_fuzzy_match"],"markdown":"## Required context\n\n- Read [/docs/llms.txt](/docs/llms.txt) first for the recommended context loading strategy.\n- Use [Agent context index](/docs/agent-context) to choose the smallest relevant context pack.\n- Use page-local markdown or llms routes for exact wording.\n- Use the manifest or page API when source confidence, source refs, related IDs, or metadata matter.\n- In an MCP client, search first and fetch the cited page or block before relying on exact wording."},{"blockId":"agent-instructions#task-intent","pageId":"agent-instructions","title":"Task intent","pageTitle":"Agent instructions","url":"https://teammately.ai/docs/agent-instructions.md","humanUrl":"https://teammately.ai/docs/agent-instructions#task-intent","markdownUrl":"https://teammately.ai/docs/agent-instructions.md","sectionId":"task-intent","kind":"reference","productArea":"reference","score":357.50946069694635,"reasons":["search_match","page_title_match","term_match","prefix_or_fuzzy_match"],"markdown":"## Task intent\n\nAnswer Teammately product and docs questions safely without inventing unsupported product claims. The agent should preserve source confidence, cite stable pages, and ask for confirmation when a claim depends on draft, inferred, or missing evidence."},{"blockId":"agent-instructions#safe-operating-boundaries","pageId":"agent-instructions","title":"Safe operating boundaries","pageTitle":"Agent instructions","url":"https://teammately.ai/docs/agent-instructions.md","humanUrl":"https://teammately.ai/docs/agent-instructions#safe-operating-boundaries","markdownUrl":"https://teammately.ai/docs/agent-instructions.md","sectionId":"safe-operating-boundaries","kind":"reference","productArea":"reference","score":356.5758221431531,"reasons":["search_match","page_title_match","term_match","prefix_or_fuzzy_match"],"markdown":"## Safe operating boundaries\n\nAgents may help readers find pages, summarize stable docs, compare source-confidence labels, and draft operational next steps. Agents should not approve policies, approve rubrics, assign reviewers, declare a candidate ready for rollout, or convert draft/inferred material into public product commitments.\n\nSearch mode is diagnostic metadata, not a confidence label. Lexical fallback remains usable when semantic embeddings are unavailable; source confidence still comes from the fetched page metadata and prose."},{"blockId":"agent-instructions#allowed-assumptions","pageId":"agent-instructions","title":"Allowed assumptions","pageTitle":"Agent instructions","url":"https://teammately.ai/docs/agent-instructions.md","humanUrl":"https://teammately.ai/docs/agent-instructions#allowed-assumptions","markdownUrl":"https://teammately.ai/docs/agent-instructions.md","sectionId":"allowed-assumptions","kind":"reference","productArea":"reference","score":346.7400300885603,"reasons":["search_match","page_title_match","term_match","prefix_or_fuzzy_match"],"markdown":"## Allowed assumptions\n\n- Code-backed pages can support object, state, and workflow claims when source_refs are present.\n- Doctrine-backed pages can support positioning, lifecycle, category language, and public narrative.\n- Docs-backed pages can support claims about routes, llms files, manifests, search, context APIs, and validation behavior.\n- Stable public pages supersede older draft or removed pages."},{"blockId":"agent-instructions#forbidden-assumptions","pageId":"agent-instructions","title":"Forbidden assumptions","pageTitle":"Agent instructions","url":"https://teammately.ai/docs/agent-instructions.md","humanUrl":"https://teammately.ai/docs/agent-instructions#forbidden-assumptions","markdownUrl":"https://teammately.ai/docs/agent-instructions.md","sectionId":"forbidden-assumptions","kind":"reference","productArea":"reference","score":346.7400300885603,"reasons":["search_match","page_title_match","term_match","prefix_or_fuzzy_match"],"markdown":"## Forbidden assumptions\n\n- Do not present draft schema pages as public schemas.\n- Do not infer public APIs, SDKs, API keys, auth behavior, rate limits, billing, compliance, tenant isolation, deployment modes, support guarantees, customer names, model-provider integrations, or production monitoring behavior.\n- Do not treat AI-assisted suggestions as approved policies, rubrics, or review context unless a stable page says a human approved them.\n- Do not collapse Teammately into eval dashboards, observability, annotation, prompt testing, generic LLMOps, or a replacement for adjacent tools.\n- Do not use internal source file names as public UI promises unless the public docs already state the behavior."},{"blockId":"agent-instructions#how-to-answer-safely","pageId":"agent-instructions","title":"How to answer safely","pageTitle":"Agent instructions","url":"https://teammately.ai/docs/agent-instructions.md","humanUrl":"https://teammately.ai/docs/agent-instructions#how-to-answer-safely","markdownUrl":"https://teammately.ai/docs/agent-instructions.md","sectionId":"how-to-answer-safely","kind":"reference","productArea":"reference","score":346.7400300885603,"reasons":["search_match","page_title_match","term_match","prefix_or_fuzzy_match"],"markdown":"## How to answer safely\n\n1. Identify whether the user is asking for positioning, operating steps, object behavior, recovery, or implementation context.\n2. Load the smallest matching context pack or query-scoped context.\n3. Check source confidence before making claims.\n4. Prefer stable page citations over broad corpus summaries.\n5. If search returns an AI Overview, verify its block references with `fetch`; treat the overview as routing assistance rather than independent authority.\n6. State uncertainty when a claim is inferred, draft-only, or outside the docs.\n7. Ask for human confirmation before advising changes that depend on permissions, compliance, deployment, billing, customer-facing APIs, or unsupported integrations."}]}