Context Retrospective
After your AI agent fumbles through a task or takes three detours to find information it should have known, run this to figure out what went wrong in your context network. It analyzes interaction transcripts across six dimensions like knowledge gaps, navigation issues, and whether the agent violated boundaries between planning docs and implementation files. The output tells you exactly what context nodes to add, what relationships to document, and which guidance rules need sharpening. Think of it as a post-mortem tool that turns every confused agent interaction into actionable improvements for your context architecture.
- skills.sh installs
- 462
- GitHub stars
- 182
- last updated
- Aug 21, 2026
Install
$ npx -y skills add jwynia/agent-skills --skill context-retrospective --agent claude-codeInstalls into .claude/skills of the current project.
Files
Context Network Retrospective
Purpose
Analyze agent-user interaction transcripts to identify context network maintenance needs and guidance improvements. Extract actionable insights for enhancing both network structure and agent instructions.
Core Principle
Learn from every interaction. Each transcript reveals gaps in context, navigation issues, and guidance problems that can be systematically fixed.
Analysis Dimensions
1. Knowledge Gap Identification
Look For:
- Questions agent couldn't answer from existing context
- Information discovered during task that should be pre-documented
- Repeated lookups of same information
- Agent confusion about structure, relationships, dependencies
Questions:
- What foundational knowledge was missing?
- Which relationships weren't documented?
- What context about history, decisions, or constraints was absent?
- Which domain boundaries were unclear?
Output: Missing information nodes and relationship gaps
2. Context Boundary Violations
Sponsors
Agent CaffeineSponsoredClose your lid, come back to finished work
SnugSponsoredThe clipboard manager for the AI eraPromote your product hereReach 125k+ visitors monthly
WaxSponsoredBlock anything with supervision, without erasing your iPhonePromote your product hereReach 125k+ visitors monthly
knobsSponsoredSee your site on any phone, any day, in any languageDetails
- Repository
- github.com/jwynia/agent-skills
- Categories
- AI & Agent Building, Design & UI/UX
- First seen
- Jun 3, 2026
More from jwynia/agent-skills
All 118- Endings461
- Governance Systems461
- Shadcn Layouts460
- Drafting454
- Story Zoom452
- Belief Systems451
- Voice Analysis451
- Reverse Outliner448
- Pwa Development447
- Chapter Drafter443
- Language Evolution439
- Book Marketing438
- Product Analysis438
- Context Network436
- Speech Adaptation436
- Competency Builder434
- World Fates432
- Sensitivity Check423
- Memetic Depth421
- Moral Parallax420
- Dna Extraction418
- Conlang417
- Gitea Workflow416
- Media Adaptation410
Recommended
More AI & Agent Building
health-check-endpointsaj-geddes/useful-ai-promptsImplement comprehensive health check endpoints for liveness, readiness, and dependency monitoring. Use when deploying to Kubernetes, implementing load balancer 461skills.sh installsserverless-architectureaj-geddes/useful-ai-promptsDesign and implement serverless applications using AWS Lambda, Azure Functions, and GCP Cloud Functions with event-driven patterns and orchestration.461skills.sh installstransaction-managementaj-geddes/useful-ai-promptsManage database transactions for data consistency. Use when implementing ACID compliance, handling concurrency, or managing transaction isolation levels.461skills.sh installswebhook-integrationaj-geddes/useful-ai-promptsImplement secure webhook systems for event-driven integrations, including signature verification, retry logic, and delivery guarantees. Use when building third-461skills.sh installs
auth0-fastapi-apiauth0/agent-skillsUse when protecting FastAPI endpoints with JWT Bearer token validation, scope checks, or DPoP binding. Integrates auth0-fastapi-api for stateless APIs receiving461skills.sh installs
endingsjwynia/agent-skillsDiagnose weak endings, rushed resolutions, and arbitrary conclusions. Use when stories build well but end disappointingly, when climax feels unearned, or when r461skills.sh installs