Answer: Hooks cannot add context before the model runs, handle delivery like replying on Slack or commenting on a PR, show PR progress, or gate a change on behavior; those should use the channel's send message and workspaceFiles/instructions.md/skills/seed files, the channel's events map or Slack and GitHub packs, githubChannel progress options, and Evals respectively. Hooks are instead used to observe every session for audit, metrics, or a transcript mirror. Status: Confirmed Entry: tool/cursor/agents-and-automation/agent-sdk-hook-boundaries Revision: 1 Digest: sha256:038835a09fa8b17d (changes whenever the content changes) Timeline: Created: 2026-11-13T03:08:53Z Last changed: 2026-11-13T03:08:53Z Last verified: 2026-11-20T12:33:49Z Next check: 2026-11-21T12:33:49Z Check interval: Daily Expected change: No fixed schedule Data: Field | Value | Source | As of Add context before the model runs | The channel's `send` message and `workspaceFiles`, `instructions.md`, skills, or `sandbox/workspace/` seed files | [1] | 2026-11-20 Delivery (Slack, PR comments) | The channel's `events` map, or the Slack and GitHub packs | [1] | 2026-11-20 Show PR progress | `githubChannel({ progress: { commitStatus, banner } })` | [1] | 2026-11-20 Gate a change on behavior | Evals | [1] | 2026-11-20 What hooks are for | A hook | [1] | 2026-11-20 Sources: [1] Anysphere. "cursor.com/docs/agent-sdk/reference/hooks.md". https://cursor.com/docs/agent-sdk/reference/hooks.md Role: primary. Accessed: 2026-11-20T12:33:49Z Quote: "Add context before the model runs | The channel's `send` message and `workspaceFiles`, `instructions.md`, skills, or `sandbox/workspace/` seed files" (Supports: Add context before the model runs) Quote: "Reply on Slack, comment on a PR, or post any other delivery | The channel's `events` map, or the Slack and GitHub packs" (Supports: Delivery (Slack, PR comments)) Quote: "Show PR progress (merge-box check, sticky banner) | `githubChannel({ progress: { commitStatus, banner } })`." (Supports: Show PR progress) Quote: "Gate a change on behavior | [Evals](https://cursor.com/docs/agent-sdk/evals.md)" (Supports: Gate a change on behavior) Quote: "Observe every session for audit, metrics, or a transcript mirror | A hook" (Supports: What hooks are for) Snapshot SHA-256: 21175b6d0a63b3713459cedd568d20c541ba9e5d98dcdc76f54c2a57951ea6a0 (captured 2026-11-13T03:08:52Z) Verification: Method: Automated extraction from [1] with deepseek-flash (JSON output, thinking disabled). Every value was checked by code to appear verbatim in a quoted passage of [1]. On this check the text of [1] was unchanged since 2026-11-13T03:08:53Z, so the values verified then were kept. Cross-check: not-available at 2026-11-20T12:33:49Z. No independent second source is configured for this entry Last successful check: 2026-11-20T12:33:49Z Last attempt: 2026-11-20T12:33:49Z Notes and limitations: Extracted automatically from a single official source (section "Hook boundaries"). Related: https://latestn.com/tool/cursor/agents-and-automation/agent-sdk-hook-events Which session events and payloads can Cursor Agent SDK hooks subscribe to? https://latestn.com/tool/cursor/agents-and-automation/agent-sdk-project-folder-structure Where do Agent SDK files go in a Cursor project? https://latestn.com/tool/cursor/agents-and-automation/agent-sdk-review-triage-templates What review and triage templates does the Cursor Agent SDK provide? https://latestn.com/tool/cursor/agents-and-automation/agent-concurrency-config How do I configure concurrency for Cursor agents in the Agent SDK? https://latestn.com/tool/cursor/agents-and-automation/agent-create-origin-repository Can a Cursor agent create an Origin repository for me? https://latestn.com/tool/cursor/agents-and-automation/agent-name-and-slug-rules How does Cursor Agent SDK determine the root agent's name and slug? https://latestn.com/tool/ All entries in AI tools Change log: Time | Event | Change | Source 2026-11-13T03:08:53Z | added | Entry created | — Changes for this entry as a feed: https://latestn.com/tool/cursor/agents-and-automation/agent-sdk-hook-boundaries/latest Challenge this entry: Open challenges: None How: If a value here is wrong or out of date, send what is wrong and a link to evidence. Challenges are queued and checked against the evidence; they do not change the entry by themselves. If the entry changes as a result, the change log says so. Form: https://latestn.com/dispute/tool/cursor/agents-and-automation/agent-sdk-hook-boundaries POST: https://latestn.com/dispute/tool/cursor/agents-and-automation/agent-sdk-hook-boundaries with claim, evidence_url, and optionally field and contact (form-encoded or JSON) GET: https://latestn.com/dispute/tool/cursor/agents-and-automation/agent-sdk-hook-boundaries?claim={what is wrong}&evidence_url={link} How to cite: LatestN. (2026). What can Cursor Agent SDK hooks not do, and what are their boundaries? Revision 1, last verified 2026-11-20T12:33:49Z. https://latestn.com/tool/cursor/agents-and-automation/agent-sdk-hook-boundaries (accessed 2026-11-21T10:03:26Z). @misc{LatestN_tool_cursor_agents_and_automation_agent_sdk_hook_boundaries_r1, title = {What can Cursor Agent SDK hooks not do, and what are their boundaries?}, author = {{LatestN}}, year = {2026}, howpublished = {\url{https://latestn.com/tool/cursor/agents-and-automation/agent-sdk-hook-boundaries}}, note = {Revision 1; digest sha256:038835a09fa8b17d; last verified 2026-11-20T12:33:49Z; accessed 2026-11-21T10:03:26Z} } Reuse: License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/) Attribution: Copying and adapting are allowed, including commercially, if you credit LatestN and link to the entry you used. Each entry ends with a ready-made citation. URL: https://latestn.com/tool/cursor/agents-and-automation/agent-sdk-hook-boundaries