aspiers/ai-config / .agents/skills/bx
bx
VerifiedCreates a Beads issue for a requested piece of work, completes exactly that work, and closes the issue. Use when the user invokes `/bx` or `$bx`, or asks for a bead to be created and the described work done in the same go. For creating a bead without doing the work, use the `bc` skill instead.
Create a bead and complete exactly the requested work
The user's request text is the sole authority for implementation scope. Do not broaden it when writing or updating the bead.
An audit, investigation, review, plan, recommendation, or proposal is complete when its requested artifact or findings are delivered. Do NOT implement, apply, or act on findings, recommendations, or proposed changes unless the request explicitly asks for that too.
Do NOT explore the codebase, launch subagents, or do any other work before creating the bead issue.
- Load the
beads-best-practicesskill and follow it when writing any bead. In particular, a bead is done by a person or by an agent, never by both: split out anything needing a human decision, approval, credential, or observation into its ownhuman-labelled bead joined by a dependency. This applies to every bead you create here, including incidental ones for follow-up work you noticed along the way. - Infer the issue type from context, using "Choose the Issue Type" in
beads-best-practices:bug,feature,chore,spike,decisionand so on. Fall back totask. - Check for an existing bead. This is the only step allowed before
creating one. Run one quick search per key title word, for example:
bd search "<word>" --status open,in_progress,blocked,deferredIf a result plainly covers the same work, use it instead of creating a new bead: tell the user, thenbd update <id> --claim. If someone else already owns it, stop and ask the user rather than working in parallel. - Otherwise create the bead and immediately mark it in progress:
bd create --title="<title>" --description="<description>" --type=<type>bd update <id> --status=in_progressUse "Investigating..." if the scope is unclear. Do not add deliverables absent from the request. - Investigate as needed, then update the bead without broadening its scope:
bd update <id> --title="<better title>" \ --description="<final description>" - Complete exactly the work requested:
- If it requests implementation or application, make the requested changes.
- If it requests only analysis, auditing, investigation, planning, review, recommendations, or proposals, produce those outputs without applying them.
- If that distinction remains ambiguous, ask the user before modifying project files.
- When the requested work, not any merely proposed follow-up, is complete,
run
bd close <id>. - If this interrupted previous work, resume it immediately without asking.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 470k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |