Gemini Precheck — Development agent for Claude Code
Pre-filter before the final Sol review.
How to install Gemini Precheck
Installs to ~/.claude/agents/juliandeans-claude-conductor-gemini-precheck.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/juliandeans/claude-conductor/HEAD/agents/gemini-precheck.md -o ~/.claude/agents/juliandeans-claude-conductor-gemini-precheck.md Restart Claude Code, or start a new session, for it to be picked up.
What Gemini Precheck does
name: gemini-precheck description: Pre-filter before the final Sol review. Has Gemini read a finished diff or subagent work to find obvious defects before Sol quota is spent on it. Worthwhile from about 400 changed lines or 8 touched files, and for mechanical or generated changes. Skip for small, careful diffs. Never replaces the Sol review. tools: Bash, Read model: sonnet effort: low maxTurns: 12 color: orange hooks: PreToolUse: - matcher: "Bash" hooks: - type: command
Alternatives in Development
- Auto Discovery — Automated agent that discovers non-obvious cross-domain connections by analyzing conceptual relationships beyo 109 ★
- Supabase Temporal Explorer — Use proactively when exploring creative, non-obvious applications of Supabase's temporal features (Realtime, Q 64 ★
- Vcf Validator — Use whenever a change touches VCF emission — adding, removing, renaming, or changing the semantics of a FORMAT 29 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Tidus
Use on a changed feature to enumerate scenarios (happy, edge, worst, abuse), mark what is already covered, wri
Quota Router
Thin wrapper over ~/.claude/router/state.sh. Reports the router's provider-state verdict (GPT/Claude/Gemini el
Verify Finding
Cheap pre-filter for §6 review findings — scores a whole batch of findings 0-100 each for whether they are rea
10x Tool Calls
Cursor and Windsurf meter usage by requests and tool calls rather than tokens, which means a finished or stall
CI Triage
Finds why CI failed and proposes the smallest fix. Use when a CI run, a make target or a hook fails and the ca
Finish Engineer
Finishing role of a topopt job. After a successful run, removes visible voxel steps and surface defects of fin
Related Skills
Claude OAuth
Run Google, xAI and OpenAI models inside Claude Code over your own OAuth sessions. The native claude path is n
Agentacct
See what your coding agents did and what it cost. Breaks each task down into work steps — tools used, files ch
Update Scaffold
Check for and apply myPKA scaffold framework updates safely. Shows a plain-English plan first (what is new, wh