Reflect Skill

Use this skill to close the feedback loop after completing a task. It reads accumulated failure signals, updates the persistent learning file, and extracts actionable rules when patterns repeat.

When to use this skill

At the end of any pipeline run that encountered failures (lint errors, test failures, MCP errors, coverage misses)
After a phase that hit its retry limit
When asked to "evaluate this session", "log what went wrong", or "update learnings"
When the Reflective Agent is not invoked as a full standalone agent but you still need to capture outcomes

Process

Step 1 — Load

Read .github/AGENT_LEARNINGS.md in full.

Note all Active Rules and their categories.
Note Failure Log entries with counts ≥ 2 (one strike away from rule extraction).

Step 2 — Evaluate & Log

Reflect Skill

Use this skill to close the feedback loop after completing a task. It reads accumulated failure signals, updates the persistent learning file, and extracts actionable rules when patterns repeat.

When to use this skill

At the end of any pipeline run that encountered failures (lint errors, test failures, MCP errors, coverage misses)
After a phase that hit its retry limit
When asked to "evaluate this session", "log what went wrong", or "update learnings"
When the Reflective Agent is not invoked as a full standalone agent but you still need to capture outcomes

Process

Step 1 — Load

Read .github/AGENT_LEARNINGS.md in full.

Note all Active Rules and their categories.
Note Failure Log entries with counts ≥ 2 (one strike away from rule extraction).

Reflect

Reflect Skill

When to use this skill

Process

Step 1 — Load

Step 2 — Evaluate & Log

Reflect

Reflect Skill

When to use this skill

Process

Step 1 — Load

Step 2 — Evaluate & Log

Step 3 — Extract Rules & Housekeeping

Output

Session Logs

OpenClaw Test Heap Leaks

Node Connect

Openclaw Qa Testing

Openclaw Secret Scanning Maintainer

Flags