Analyzes recent OpenCode sessions to recommend reusable skills, agents, commands, and config changes.
npx skills add https://github.com/Favot/.dotfiles --skill reflect-favot
Please help me install this Agent Skill. Skill: reflect Source: https://github.com/Favot/.dotfiles/tree/main/.config/opencode/skills/reflect Command: npx skills add https://github.com/Favot/.dotfiles --skill reflect-favot
--sessions, queries the OpenCode SQLite database to analyze historical sessions across repos, caching per-session summaries and aggregating cross-repo patterns. - Smallest-Useful-Form Recommendations: Proposes the least powerful fix (prompt rule, skill, command, agent, playbook, or no change) and asks for approval before editing any files. - Use Case: After noticing you repeatedly run the same release checklist, run /reflect release workflow to get an evidence-backed proposal for a reusable command or playbook. ## Quick Start Ask the agent to run /reflect on your recent work to identify repeated workflows worth turning into reusable assets.High-intent search queries and answers about installing and using this skill.
Run /reflect --sessions to query the OpenCode SQLite database at ~/.local/share/opencode/opencode.db. It loads recent sessions and messages, summarizes each with friction points and confidence scores, then aggregates cross-repo patterns into a report.
Reflect scores each candidate on frequency, cost, risk, stability, and existing coverage, then recommends the smallest useful form. Prompt rules come first, then skills, commands, and only custom agents when a distinct specialist lane is clearly needed.
No. Reflect presents a proposal and asks for approval before editing prompts, skills, commands, agents, MCP permissions, or config. When changes are applied, they take effect on the next OpenCode run, so a restart may be needed.
It reports that no strong repeated workflow was found and recommends no changes. Treating 'create nothing' as a valid outcome is part of its conservative, evidence-driven contract.
Sessions are skipped when they fail to load or contain no messages. Reflect logs each failure, continues with remaining sessions, and reports the total skipped count at the end of the analysis.
korchasaAnalyze past IDE session history to identify recurring patterns and propose improvements.
MDSIXONEReviews recent sessions to recommend reusable skills, agents, commands, or config changes.
mattbutlerengineeringConfigure OpenCode AI coding assistant repository settings via opencode.json.
korchasaAnalyze historical IDE session transcripts to classify recurring agent behavior patterns and issues.