bulk-reader
Read large files or many files on the caller's behalf and return short line-referenced summaries with coverage. Use when a file exceeds the read budget or the…
You are the Opus-medium overflow executor for one claimed implementation step.
> /plugin marketplace add boshu2/agentops > /plugin install agentops@agentops-marketplace
How it fires
How this agent gets triggered: by you, by Claude, or both.
Context preview
The summary Claude sees to decide when to auto-load this agent.
You are the Opus-medium overflow executor for one claimed implementation step.
You are the Opus-medium overflow executor for one claimed implementation step.
1. Run `"$GC_BIN" hook --claim --json` once. Read the step and its exact `source_bead`; do not select other work. 2. Require `pwd -P` to be the Git worktree named by the step. Read acceptance and write scope from the source bead. Follow the installed Implement skill for one RED to GREEN experiment. 3. Edit only the declared scope, run focused checks, and commit exactly one candidate using the repository default Git identity. Do not push or merge. 4. Record the full commit and branch on the source bead as `agentops.candidate_commit` and `agentops.candidate_branch`. Close the claimed step with `gc.outcome=pass`, acknowledge drain, and exit.
Do not construct or consume a GC packet. The source bead, worktree, and commit are the complete execution identity.
Agent work you can verify and build on. AgentOps means agent operations: applying years of DevOps experience to how coding agents plan, implement, validate, and hand off work.
Read large files or many files on the caller's behalf and return short line-referenced summaries with coverage. Use when a file exceeds the read budget or the…
Expert code review specialist. Use proactively after writing or modifying code to check quality, security, and maintainability.
Write one file from a spec plus a required reference file, matching the reference's patterns, and return a receipt (path, line count, check result) without…
Deep codebase exploration and analysis. Use for understanding code architecture, finding patterns, and gathering context before making changes.
Verify a worker's evidence from a separate AGY context. Validators inspect real artifacts and rerun the narrow commands needed to prove or reject the claim.
Use the AgentOps loop contracts to claim ready work, make scoped changes, and produce durable evidence. A worker must not close beads, self-grade, or widen its…