Better Every Run

Other

Better Every Run: turn user corrections into preferred future outcomes through a simple /ber-style fix, remember, and report flow.

Install

openclaw skills install better-every-run

Better Every Run

Use this skill when the user corrects an outcome, states a preference, asks why something keeps failing, or wants an agent to improve future behavior from the current run.

The human path is deliberately small:

/ber fix bad outcome -> desired outcome
/ber remember simple rule
/ber report

The agent runs the bundled local helper in the background and reports the result in chat.

When To Use

  • The user corrects an agent mistake or preference.
  • A tool/workflow failed in a way likely to repeat.
  • A workflow succeeded and should become reusable.
  • The user says "remember this", "next time", "always", or "never".
  • Before ending a substantial session where useful lessons were found.

Human Commands

/ber fix agent wrote vague status -> agent gives exact command output and next action
/ber remember use the approved development host for active code work
/ber report

Rules

  • Report CLI output back in chat; do not build web pages or dashboards.
  • Do not silently edit MEMORY.md, AGENTS.md, SOUL.md, or other durable instruction files.
  • Keep corrections factual: bad outcome, desired outcome, target file if applying.
  • Avoid private data unless the user explicitly wants it captured.
  • Hide helper complexity from the user; never make them manage accept/export/apply steps during normal use.
  • Design for the shortest path to the user's outcome.

Workflow

For normal chat use, keep the visible flow to one command:

/ber fix bad outcome -> desired outcome

For simple one-line preferences:

/ber remember simple rule

Use /ber report to show what was learned.

Agent Implementation

Keep helper paths, target-file selection, ledger files, and memory-patch mechanics out of the human-facing answer. The agent may use them internally for deterministic execution, tests, and audits.

Read references/workflow.md for the full workflow. Read references/report-template.md for expected chat output shape.