OpenSkillIndex← back to search
claude-code

GAAI-framework post-mortem-learning

Analyze failures and suboptimal deliveries to identify root causes, contributing factors, and raw lessons. Activate after significant delivery failures, repeated QA failures, or when patterns of issues need to be understood.

install
source · Clone the upstream repo
git clone https://github.com/Fr-e-d/GAAI-framework
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/Fr-e-d/GAAI-framework "$T" && mkdir -p ~/.claude/skills && cp -r "$T/.gaai/core/skills/cross/post-mortem-learning" ~/.claude/skills/fr-e-d-gaai-framework-post-mortem-learning && rm -rf "$T"
manifest: .gaai/core/skills/cross/post-mortem-learning/SKILL.md
source content

Post-Mortem Learning

Purpose / When to Activate

Activate after:

  • Significant delivery failures
  • Repeated QA failures on the same area
  • Delivery that missed its success metrics
  • Patterns of issues that need systemic understanding

Process

  1. Reconstruct what happened end-to-end
  2. Compare expected vs actual outcomes
  3. Identify: technical causes, contextual gaps, decision errors, rule weaknesses
  4. Produce concrete failure narratives
  5. Extract raw lessons (non-generalized) — specific to this failure

Outputs

  • Root cause analysis report
  • Failure timeline
  • Mapped rule gaps
  • Raw lesson list (specific, not generic)
  • Candidate improvement areas (for backlog or rules)

Quality Checks

  • Root causes are specific, not generic ("insufficient testing" is invalid — "acceptance criterion #3 had no test coverage" is valid)
  • Lessons are raw and specific, not platitudes
  • Failure timeline is traceable
  • Rule gaps link to specific rule files

Non-Goals

This skill must NOT:

  • Propose solutions (produces inputs for backlog and rules, not fixes)
  • Assign blame
  • Generalize into vague lessons

Makes failures understandable and actionable. Enables serious continuous improvement.