$npx -y skills add Codagent-AI/agent-skills --skill handoffSummarizes a specific aspect of the current conversation so another agent can resume. Use when the user says "handoff", "summarize for handoff", "write a handoff", or wants to capture context for another agent session. The user MUST specify what aspect to summarize in their promp
| 1 | # Handoff |
| 2 | |
| 3 | Write a focused summary of a specific aspect of the current conversation so a fresh agent can pick up where this session left off. |
| 4 | |
| 5 | This is NOT a full conversation summary. The user's prompt tells you which aspect to capture. If the aspect is unclear, ask for clarification before proceeding. |
| 6 | |
| 7 | ## What to capture |
| 8 | |
| 9 | For the aspect the user specified, write: |
| 10 | |
| 11 | 1. **Objective** — What the user is trying to accomplish (the goal, not the history). |
| 12 | 2. **Current state** — Where things stand right now. What's done, what's in progress, what's blocked. |
| 13 | 3. **Key decisions** — Decisions made during this session that the next agent must honor, with brief rationale. |
| 14 | 4. **Open questions** — Unresolved issues, trade-offs still being weighed, or things the user hasn't decided yet. |
| 15 | 5. **Next steps** — Concrete actions the resuming agent should take first. |
| 16 | 6. **Relevant files** — Paths to files that were created, modified, or are central to the work. |
| 17 | |
| 18 | Omit any section that has nothing meaningful to say. Keep it tight — the resuming agent needs signal, not narration. |
| 19 | |
| 20 | ## Output |
| 21 | |
| 22 | Write the summary to `handoff.md`. |
| 23 | |
| 24 | **Determining the output location:** |
| 25 | |
| 26 | 1. If the user specified a path in their prompt, use it. |
| 27 | 2. If a project configuration (e.g., AGENTS.md, openspec config, or similar conventions) specifies a change directory, write there. |
| 28 | 3. Otherwise, determine the relevant change slug from context (the change being worked on, or a kebab-case slug from the topic) and write to `~/.agent-skills/changes/<slug>/handoff.md`. |
| 29 | |
| 30 | Do not ask the user to confirm the location. Just write to the resolved path. |
| 31 | |
| 32 | ## Template |
| 33 | |
| 34 | ```markdown |
| 35 | # Handoff: <aspect being handed off> |
| 36 | |
| 37 | ## Objective |
| 38 | |
| 39 | <!-- What the user is trying to accomplish --> |
| 40 | |
| 41 | ## Current State |
| 42 | |
| 43 | <!-- Where things stand: done, in progress, blocked --> |
| 44 | |
| 45 | ## Key Decisions |
| 46 | |
| 47 | <!-- Decisions made this session that the next agent must honor --> |
| 48 | - **<decision>** — <rationale> |
| 49 | |
| 50 | ## Open Questions |
| 51 | |
| 52 | <!-- Unresolved issues or trade-offs --> |
| 53 | |
| 54 | ## Next Steps |
| 55 | |
| 56 | <!-- Concrete first actions for the resuming agent --> |
| 57 | |
| 58 | ## Relevant Files |
| 59 | |
| 60 | <!-- Paths to files central to this work --> |
| 61 | ``` |
| 62 | |
| 63 | ## Guardrails |
| 64 | |
| 65 | - **Write from current context only.** The handoff should be based on what the agent already knows from the current conversation — do not explore files, run searches, or do additional research unless the user explicitly asks for it. |
| 66 | - **Do not summarize the entire conversation.** Focus only on the aspect the user specified. |
| 67 | - **Do not invent context.** Only capture what actually happened or was discussed in this session. |
| 68 | - **Do not prescribe solutions for open questions.** State the question; let the next agent and user decide. |
| 69 | - **Write for a cold reader.** The resuming agent has zero prior context — be precise about file paths, branch names, and terminology. |