From amplify
Recovers from repeated tool call errors, unexpected failures, or blockers during plan execution by checking the plan, verifying alignment, and fixing specific issues without drifting.
How this skill is triggered — by the user, by Claude, or both
Slash command
/amplify:recover-from-errorsThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
When you encounter repeated errors with tool calls (e.g., wrong arguments, missing files, permission issues), **DO NOT** simply generalize or guess new parameters, alternative tool locations, as this often leads to drifting away from the original goal by introducing noise to the context.
When you encounter repeated errors with tool calls (e.g., wrong arguments, missing files, permission issues), DO NOT simply generalize or guess new parameters, alternative tool locations, as this often leads to drifting away from the original goal by introducing noise to the context.
Check the Plan: Immediately look up the plan file of the session to verify the context of the current operation.
Verify Alignment: Determine if the tool call that generated the error is actually part of the tasks currently being executed in the plan. Ask Yourself: "Is the action I just attempted explicitly required by the current step in the plan?"
Re-align to the Plan:
npx claudepluginhub wezzard/skills --plugin amplifyUse when repeated fix attempts fail, the agent appears stuck in a loop, or complexity is increasing without progress
Resolves conflicting demands, tool failures, and scope creep by blending with pressure then reframing toward productive resolution.
Guides structured recovery from errors and failures by assessing, logging root cause, and applying safe fixes with rollback plans. Prevents cascading mistakes via forbidden patterns and escalation rules.