agent-task-handoff
Agent Task Handoff
When delegating work to a sub-agent or across sessions, make the handoff self-contained.
Handoff package must include:
- Goal: What the sub-agent must produce (one clear sentence).
- Context: The minimum background the agent needs — no more, no less.
- Constraints: Explicit rules (don't modify X, use library Y, output format Z).
- Inputs: File paths, API endpoints, or data the agent needs to start.
- Success criteria: How to verify the result is correct.
For parallel agents:
- Clearly state which parts of the codebase/data each agent owns.
- Define how outputs will be merged back.
Anti-pattern: Sending a vague "continue where we left off" with no context — sub-agents start from zero.
More from aiming-lab/metaclaw
structured-step-by-step-reasoning
Use this skill for any problem that involves multiple steps, tradeoffs, or non-trivial logic. Think out loud before answering to improve accuracy and transparency. Apply whenever the answer is not immediately obvious.
14avoid-hallucinating-specifics
Common mistake — stating specific facts (API endpoints, library versions, config options, function signatures) with false confidence when uncertain. Always flag uncertainty rather than guessing specifics.
14codebase-navigation
Use this skill when exploring an unfamiliar codebase, tracing code paths, or answering questions about how the system works. Read before writing, and build a mental model of the architecture before making changes.
12graceful-error-recovery
Use this skill when a tool call, command, or API request fails. Diagnose the root cause systematically before retrying or changing approach. Do not retry the same failing call without first understanding why it failed.
11uncertainty-acknowledgment
Use this skill when you are not sure about a fact, have outdated knowledge, or the question is contested. Explicitly communicate the level of confidence instead of asserting uncertain things as fact.
11secure-code-review
Use this skill when reviewing or writing code that handles user input, authentication, file I/O, network requests, or database queries. Always check for common security vulnerabilities before considering the code complete.
10