Stats
Actions
Tags
Help us improve
Share bugs, ideas, or general feedback.
From context-files
Review recent sessions for the current project to identify recurring patterns, systemic issues, and improvements across multiple sessions.
npx claudepluginhub maxnoller/claude-code-plugins --plugin context-filesHow this skill is triggered — by the user, by Claude, or both
Slash command
/context-files:sessions-reviewThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Session ID: ${CLAUDE_SESSION_ID}
Reviews recent Claude Code sessions (default last 5) in the current project for patterns, using parallel conversation-reviewer agents per session and cross-session synthesis.
Root cause analyzer for AI coding sessions. Detects scope drift, rework patterns, and root causes to improve future prompts and repo health.
Analyzes Claude Code session logs to extract tool usage stats, thinking blocks, error patterns, debug trajectories, and generate actionable productivity recommendations. Provides cc-session CLI for overviews, timelines, searches.
Share bugs, ideas, or general feedback.
Session ID: ${CLAUDE_SESSION_ID}
Use Glob to locate the current session file:
~/.claude/projects/**/${CLAUDE_SESSION_ID}.jsonl
Take its parent directory and glob all *.jsonl files there — these are all sessions for this project. Sort by modification time and take the 10 most recent (excluding the current session).
For each session file, use Grep to extract conversation turns ("type":"user" and "type":"assistant" lines) to reconstruct what happened. Note the timestamp and approximate length of each session.
Look for patterns that appear across multiple sessions — single-session issues belong in a regular /session-review. Focus on:
Recurring mistakes — the same wrong assumption or error appearing in 2+ sessions. Strong signal that something needs fixing in the codebase, tooling, or context file.
Persistent friction — the same awkward workflow showing up repeatedly. A script or hook could likely automate it.
Knowledge gaps — things the agent keeps having to look up or ask about that should be self-evident from the code.
Drift — something that worked well in early sessions but degraded later, or vice versa.
What's consistently smooth — patterns worth preserving.
For each finding, assign exactly one of:
Output format:
Sessions analyzed: N (date range)
Recurring findings (seen in 2+ sessions, prioritized by frequency and impact):
What's consistently smooth (brief)
A pattern appearing once is noise. Appearing twice is a coincidence. Appearing three times is a system problem — fix the system, not the symptoms.