Salberg87

Salberg87/batonq-skills

Plugin Claude Code

Multi-CLI role skills for batonq autonomous coding agents — worker, judge, pr-runner, explorer, reviewer.

Salberg87/batonq-skills

Instructions file CodexOpenCode

AGENTS.md instructions for Salberg87/batonq-skills, a project described as: Multi-CLI role skills for batonq autonomous coding agents — worker, judge, pr-runner, explorer, reviewer.

Salberg87/batonq-skills

Instructions file

Claude Code instructions for Salberg87/batonq-skills: This is the curl-installable mirror for the batonq-worker skill — when this file is dropped into a project root, Claude Code picks it up as the default behavior for any agent working in that project.

Salberg87/batonq-skills

Skill Claude CodeCodex

Read-only codebase reconnaissance. Maps the relevant files, functions, and dependencies for an upcoming task without making any edits. Output is a brief written report. Use as a cheap first pass (haiku, flash) before dispatching a worker — surfaces gotchas the worker would otherwise discover late, when changing…

Salberg87/batonq-skills

Skill Claude CodeCodex

Reviews a git diff against a task spec and outputs a PASS/FAIL verdict. Output is verdict-only — no preamble, no remediation suggestions, no narrative. Use when running batonq's judge gate as a fresh agent invocation, where you receive (a) the original task body, (b) the verifycmd output, (c) the git diff since claim…

Salberg87/batonq-skills

Skill Claude CodeCodex

Implements a batonq task end-to-end and opens a pull request. Adds git discipline, branch naming, conventional-commit enforcement, and PR-body conventions on top of batonq-worker. Use when a task explicitly requires a PR (not a direct commit to main) — e.g., tasks running in fire-and-forget mode against repos with…

Salberg87/batonq-skills

Skill Claude CodeCodex

Reviews code from another agent (or another CLI tool) without knowing which one wrote it. Outputs a ranked critique using stable letter labels (A/B/C) instead of model or tool names. Anti-bias by construction. Use in cross-tool eval flows where the same task was solved by N different runners and you want a per-trait…

Salberg87/batonq-skills

Skill Claude CodeCodex

Default coding agent for tasks picked from a batonq queue. Karpathy-style discipline (think before coding, simplicity, surgical changes, verifiable goals) plus batonq-specific gates (verify and judge are mandatory; no skip flags). Use when implementing a normal task — fixing a bug, adding a feature, refactoring code.