trackbed:code:audit-practice
Check how you work with coding agents against 18 common mistakes, from effort and alignment to review and steering, and get the fixes to make first.
/trackbed:code:audit-practice [--recent N] [--no-sessions]- Invocation: User
- Hosts: Claude Code, Codex (
$trackbed:code:audit-practicein Codex) - Category: Audits
- Source:
trackbed/skills/code:audit-practice/SKILL.md
Description
Section titled “Description”Audits agentic coding practice against eighteen common mistakes, from effort level, skill length, and alignment to review balance, self-verification, babysitting, global steering files, codebase navigation, and strategic thinking. Gathers repository, agent-config, and recent-session evidence and replies inline with the fixes to make first. Read-only. Not for per-entry config audits (trackbed:code:audit-agent-config) or agent-readiness scores (trackbed:code:agent-readiness).
Related
Section titled “Related”- Refers to:
trackbed:code:agent-readiness,trackbed:code:audit-agent-config,trackbed:code:plan-to-pr,trackbed:code:refactor-opportunities,trackbed:discovery:interview,trackbed:discovery:strategic-inquiry,trackbed:discovery:wayfinder,trackbed:docs:adr,trackbed:docs:feature-spec,trackbed:docs:ubiquitous-language,trackbed:docs:update-agents-md,trackbed:learn:verify-understanding,trackbed:linear:issue-define,trackbed:linear:issue-to-pr,trackbed:pr:adversarial-review,trackbed:pr:code-review,trackbed:pr:review-convergence,trackbed:pr:triage-risk,trackbed:product:strategy,trackbed:qa,trackbed:session:automate-repeats,trackbed:skill:review,trackbed:test:audit,trackbed:verify:run