Retrospect Collab

Trust Score 85/100

Analyze collaboration and human–AI interaction across sessions to surface metrics, patterns, and concrete Start/Stop/Continue recommendations.

triggers:retrospect collabcollaboration analysissession patternshow am I collaboratingretrospective review

GitHub SKILL.md

What it does

This skill analyzes captured session logs to measure collaboration quality across two dimensions: technical effectiveness (context management, guidance, critical thinking, bias) and cognitive posture (intentionality, agency, impact). It extracts prompts, tool calls, durations, and events to produce quantified summaries and actionable Start/Stop/Continue recommendations written to a retro insights file.

When to use it

Use when you want to audit a set of sessions (team retros, agent runs, human–AI experiments) over a time period. Typical triggers: retrospective reviews, periodic QA of agent interactions, or when tracking improvement in collaboration skills across releases.

What's included

Scripts: none embedded in the skill file (has_scripts=false) but the repo includes reference.md with templates.
References: reference.md (scoring criteria and report templates).
Instructions: step-by-step filtering, extraction, scoring, and reporting guidance; instructions show how to load sessions and write insights to .retro/insights/collab/{PERIOD}.md.

Compatible agents

Best for agents and workflows that can access session logs and run simple shell helpers (Cursor/Claude plugin-style environments); works as a guidance skill for evaluation rather than a UI tool.

Audit Summary

Retrospect Collab is a collaboration analysis skill that guides agents through analyzing human-AI session patterns across technical effectiveness and cognitive posture dimensions. The SKILL.md is well-written with clear steps, role definition, and evidence-based principles, but references an external script (retrospect-load-sessions.sh) and reference.md that aren't bundled, making it incomplete and non-functional out of the box. No security concerns — purely instructional with no executable code.

Watch Out

References scripts/retrospect-load-sessions.sh and reference.md that aren't bundled — skill won't work without them
Depends on CLAUDE_PLUGIN_ROOT env var which is Claude-specific, not portable across agent platforms
No fallback instructions if sessions directory is empty or doesn't exist

Notes

Well-structured skill conceptually with good analytical framework (Start/Stop/Continue, impact categorization, bias awareness). The cynefin-domain and verb annotations are a nice touch. Main weakness is lack of bundled resources making it non-functional without the original repo.

Information

Repository: agent-skills
Stars: 13

Trust Score

Overall85

Security100

Code Quality72

Architecture68

Usefulness52

More from agent-skills

Donella Meadows Philosophical Dialogue

Instantiates a high-fidelity persona of Donella Meadows for deep systems thinking, sustainability, and leverage point analysis.

Related Skills

Weekly Retro (Personal Corp)

Run a structured weekly retrospective: gather facts from repos and tools, interview the founder, and turn decisions into canonical files and issues.

Skill Retro

Analyze a session's skill invocations to surface performance issues and recommend or apply improvements using the skill-creator workflow.

AI News Daily

Generate a structured, bilingual daily briefing of the latest AI industry news, model updates, papers, and open-source tools with in-depth technical analysis an

Analyzing Ransomware Leak Site Intelligence

Collect and analyze ransomware data-leak site (DLS) posts to extract victim, group, sector and geographic trends for threat intelligence and proactive defense.

Analyzing Ransomware Leak Site Intelligence

Collect and analyze ransomware data-leak site (DLS) postings to extract victim, group, sector, and timeline intelligence for threat hunting and risk assessment.

5S Auditor

Conduct structured 5S workplace audits with scoring, photo documentation, trend analysis, and action-item tracking to improve shop-floor organization and sustai

Arcana (Technical Mechanism Analysis)

Explain how systems, protocols, or architectures actually work under the hood and surface non-obvious implications and failure modes.

PESTEL Analysis (Product Management)

Run a structured PESTEL analysis to identify political, economic, social, technological, environmental, and legal forces that affect product strategy and market

Back to Skills