2.0 KiB
2.0 KiB
name, description, compatibility
| name | description | compatibility |
|---|---|---|
| counselors | Run multi-agent code reviews with the counselors CLI and synthesize actionable findings. | opencode |
Counselors Skill
Use this skill when the user asks for second opinions, multi-model review, architecture critique, or bug/risk hunting.
Core workflow
-
Gather context quickly.
- Identify the target files and question scope.
- If needed, inspect
git diff HEADandgit diff --staged.
-
Discover configured tools and groups.
- Run:
counselors lscounselors groups ls
- Show the full outputs to the user and ask which tools/group to use.
- Run:
-
Build prompt file with counselors.
- Use
counselors mkdir --jsonand pipe prompt content in. - Use
@path/to/filereferences instead of inlining large files.
- Use
-
Dispatch run.
- Single pass:
counselors run -f <promptFilePath> --tools <ids> --read-only best-effort --json - Iterative pass:
counselors loop -f <promptFilePath> --tools <ids> --read-only best-effort --json - Preset loop:
counselors loop --preset <preset> "<focus>" --tools <ids> --read-only best-effort --json
- Single pass:
-
Read outputs and synthesize.
- Parse the JSON manifest.
- Read each
outputFilefor successful tools. - Check
stderrFilefor failed/empty tools. - Produce concise synthesis:
- consensus
- disagreements
- key risks
- recommendation
- Include the saved output directory path.
Usage guardrails
- Ask for tool/group selection before dispatch.
- Confirm final tool list before running.
- Prefer
--jsonoutput for parseability. - Use long timeouts for dispatch (often 10+ minutes).
- If selected tools include custom
opencode-*, avoid--read-only strictbecause those tools are configured asbestEffortand will be skipped under strict policy. - If no tools are configured, instruct user to run
counselors init --auto.
Useful commands
counselors init --autocounselors tools addcounselors lscounselors groups lscounselors run --helpcounselors loop --help