bulma-goal
Skill1 starsChanged 50 days agoMITView on GitHub
Drive developer → CI green → QA comment+evidence across turns with schedule_wake. Use when /goal or /bulma-goal, or after a draft PR is opened and CI is still pending.
In this repository35 other files it gives its agents.
- AGENTS.mdAGENTS.md
- skills/bulma-feature-delivery/adapters/claude.mdCLAUDE.md
- architectSkill
- before-and-afterSkill
- blast-radiusSkill
- bulma-busSkill
- bulma-code-reviewSkill
- bulma-developerSkill
- bulma-fdSkill
- bulma-feature-deliverySkill
- bulma-hostSkill
- bulma-lstmSkill
- bulma-memorySkill
- bulma-qaSkill
- bulma-reviewerSkill
- bulmaSkill
- bulma-triageSkill
- diagnoseSkill
- figure-it-outSkill
- grill-with-docsSkill
- howSkill
- interrogateSkill
- kevinSkill
- no-commentsSkill
- principle-fix-root-causesSkill
- principle-prove-it-worksSkill
- principle-sequence-verifiable-unitsSkill
- receiving-code-reviewSkill
- tddSkill
- technical-writingSkill
- thermo-nuclear-code-quality-reviewSkill
- to-issuesSkill
- to-prdSkill
- typescript-best-practicesSkill
- unslopSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 1 is distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 44
If the host has a /goal (or equivalent) command, register that sentence there.
- Very specific: line 47
If there is no PR yet → load bulma-developer GRAPH (deliver).
- Very specific: line 48
When a draft PR exists and CI is not green → schedule_wake (LOOP.md).
- Distinctive: line 50
Stop the turn. The wake resumes this skill.
- Very specific: line 54
Read STATE. Inspect the real PR (gh pr view, gh pr checks, issue comments).
- Very specific: line 75
Start a second QA while qa_active holds another live PR
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Bulma goal / loop
The graphs do not finish in one turn. CI takes 20–30 min. This skill keeps the session working until QA has commented with evidence.
REQUIRED: COMPLETE.md ·
LOOP.md · bus PROTOCOL.md ·
DISK.md (.bulma-* is global, never git root) ·
bulma-memory
Chat: bulma-memory. Apply Kevin for human-facing communication.
Worktree and branch rules: JOBS.md §Worktree.
Commands
| Args | Action |
| ticket id / feature text | Start goal + developer graph |
PR url / owner/repo#N | Start goal on that PR (skip implement if it exists) |
status | Read .bulma-goal/STATE.md + gh pr checks + last QA comment |
cancel | Delete the scheduler loop; leave graphs as-is |
Start
- Load
bulma-memory. Write.bulma-goal/STATE.mdfrom templates/STATE.md. - Name the completion bar (verifier / user):
Draft PR for <id> is CI-green, MERGEABLE, and has a bulma-qa
comment on the head SHA that includes evidence.If the host has a /goal (or equivalent) command, register that
sentence there. If it does not, STATE is enough.
- If there is no PR yet → load bulma-developer GRAPH (
deliver). - When a draft PR exists and CI is not green → schedule_wake
(LOOP.md). Do not
gh pr checks --watch. - Stop the turn. The wake resumes this skill.
Each wake (or /bulma-goal resume)
Read STATE. Inspect the real PR (gh pr view, gh pr checks,
issue comments). Then one step:
| World | Next |
| No PR | developer deliver |
| CI pending / in progress | stop (wait for next fire) |
| CI red | developer fix / fd ci loop |
| Green but not MERGEABLE | developer mergeable |
| Green + MERGEABLE, no QA comment on this SHA | enqueue-or-start QA (JOBS.md) |
| QA comment on this SHA, verdict FAIL + PR_BUG | developer fix |
| QA comment on this SHA, PASS / FAIL-unrelated / BLOCKED documented | complete |
Complete = COMPLETE.md all true, then
cancel_wake.
Never
- Claim the goal done without the QA comment + evidence on head SHA
- Sit in
--watch(CI is longer than a tool timeout) - Spawn graph-agents as children (bus switch / load GRAPH)
- Start a second QA while
qa_activeholds another live PR - Merge
After QA comment
Cancel the loop. Report in the chat language: PR, SHA, QA verdict, comment URL, evidence URL.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.