agentleFS
adezdevadezdev/blueprint/skills/build/SKILL.md

build

Skill0 starsChanged 3 months agoMITView on GitHub

Execute the next unblocked task from a Blueprint tasks.md — read only that task's component and acceptance criteria, make the focused change, prove it with the done-when check, tick the box. Use when the user says "/blueprint:build", "do the next task", "implement T4", "continue the blueprint", or wants the spec turned into code one reviewable commit at a time.

  • Commits and pushes

156 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use backprop, finances and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • JuliusBrusseeSkill

    backprop

    Bug → spec protocol. When a bug is found or a test fails, trace the cause, decide whether a new §V invariant would catch recurrence, append to §B. This is the one non-obvious thing SDD does that plan-then-execute doesn't. Triggers on test failure, bug report, post-mortem, or explicit user ask.

    Also in 4 projects that use this

  • GerardoRdz96Skill

    finances

    Use when Gera talks about his money — logging an expense ("I spent X on Y"), checking where he stands this pay-period, closing a period, planning debt payoff, onboarding his financial picture, or asking for budgeting/finance opinions. Servy's financial brain. Triggers — "I spent", "log this expense", "where am I this period", "close the period", "how's my budget", "debt plan", "should I pay off", "finances onboard", "/finances".

    Also in 3 projects that use this

  • cpiprintSkill

    favicon

    Generate favicons from a source image

    Also in 3 projects that use this

  • madeehahassanSkill

    Go-Live

    Pre-launch checklist. Use before making any app public, sharing a link externally, or announcing a product launch. Adapts to Claude Code, claude.ai, and Cowork.

    Also in 3 projects that use this

  • JuliusBrusseeSkill

    deepen

    Optional design-improvement pass for when you have spare usage to drain. Finds the shallowest modules in the code the spec touches, researches a deeper design, and proposes refactors that shrink interfaces and hide decisions — behavior held constant, tests green before and after. Proposes §I/§V/§T edits, never silent rewrites. Triggers when the user says "deepen this", "improve the design", "this module feels shallow", "pull complexity down", "use spare budget on the codebase", or invokes /ck:deepen. Leans on the codebase-design skill's deep-module vocabulary when present.

    Also in 3 projects that use this

  • synaptiaiSkill

    gate

    Create a BDSK review_gate artifact. Defines a checkpoint that controls whether generated output may proceed toward acceptance. Use when defining quality, security, or scope gates for execution plans.

    Also in 5 projects that use this

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. 0 are common, 8 are distinctive, and 6 are very specific to this project.

  • 0

    Common

    Most of the 143 edited versions we read keep these.

  • 8

    Distinctive

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read ${CLAUDE_PLUGIN_ROOT}/FORMAT.md first (or FORMAT.md at the repo root if you are working inside the blueprint repo itself).

    line 8
  2. Distinctive:

    All tasks checked → say so, and go to Phase close below.

    line 23
  3. Distinctive:

    Do not read the whole spec, and do not implement anything that is not in the cited component.

    line 35
  4. Very specific:

    If on the repository's default branch, create a feature branch first (blueprint/<slug>).

    line 41
  5. Very specific:

    Write the test that asserts the cited ACs, in the same pass.

    line 47
  6. Very specific:

    Run done-when. Paste the real output.

    line 50
  7. Very specific:

    Only after done-when genuinely passes, tick the box in tasks.md: - [x] T7 ....

    line 53
  8. Distinctive:

    Do not commit unless the user asked.

    line 56
  9. Distinctive:

    Stop and ask if, while implementing, you find that:

    line 61
  10. Distinctive:

    When the last task in a phase is ticked:

    line 75
  11. Distinctive:

    Run the full test suite, not just this task's check.

    line 77
  12. Very specific:

    Run /code-review on the branch diff.

    line 78
  13. Distinctive:

    Never tick a box without a passing check.

    line 99
  14. Distinctive:

    Never edit requirements.md or architecture.md from this skill.

    line 100

“Common” is measured: kept by at least 6 in 10 of the 143 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Blueprint: build

Read ${CLAUDE_PLUGIN_ROOT}/FORMAT.md first (or FORMAT.md at the repo root if you are working inside the blueprint repo itself).

One task per invocation. Stop when it is done. That boundary is the point of this skill — it is what keeps a diff reviewable.

Select the task

Argument may be a slug, a task id (T4), both, or nothing.

  1. Resolve the feature: argument slug → single .blueprint/*/tasks.md → ask.
  2. Explicit T<n> given → use it, but refuse if its (after ...) dependencies are unchecked. Say which.
  3. Otherwise take the first [ ] task, in phase order, whose dependencies are all [x].
  4. All tasks checked → say so, and go to Phase close below.

Load only what the task cites

This is the discipline that makes the change focused. For task T7 ... → C2 | R1.AC1, R1.AC2, read:

  • the ### C2 section of architecture.md — files, interface, dependencies
  • the R1.AC1 and R1.AC2 lines of requirements.md
  • the files the task names, and their existing tests
  • whatever those files actually call, as far as you need to understand them

Do not read the whole spec, and do not implement anything that is not in the cited component. Work you notice is needed but that no task covers is a finding, not a licence — note it and raise it at the end.

Execute

  1. If on the repository's default branch, create a feature branch first (blueprint/<slug>). Do not commit to main directly.
  2. Make the change. Match the surrounding code — its naming, its error handling, its test style. The component interface in architecture.md is a contract; if you must deviate from the declared signature, that is a spec change, see Escalate.
  3. Write the test that asserts the cited ACs, in the same pass. A task whose done-when is a test command is not done until that test exists and fails for the right reason before it passes.
  4. Run done-when. Paste the real output. If it fails, fix and rerun. If it fails for a reason outside this task's scope, stop and report — do not widen the diff to make a check go green.
  5. Only after done-when genuinely passes, tick the box in tasks.md: - [x] **T7** .... Never tick a box on the strength of "the code looks right".
  6. Do not commit unless the user asked. Report what changed and let them look.

Escalate, do not improvise

Stop and ask if, while implementing, you find that:

  • the component's declared interface does not work
  • the task depends on something no earlier task built
  • an acceptance criterion is untestable or contradicts another
  • the change needs files outside the component's declared set

Say which document is wrong and what the amendment would be (/blueprint:architecture <slug> to add C7, or a new R-id). Silently deviating from the spec destroys the traceability that the rest of this plugin depends on — the docs stop describing the code and become a lie.

Phase close

When the last task in a phase is ticked:

  1. Run the full test suite, not just this task's check.
  2. Run /code-review on the branch diff. Report the findings; apply only what the user approves.
  3. Offer a PR. If accepted, gh pr create with a body that lists the phase's task ids and the requirements they satisfy, so a reviewer can check the claim:

    ## Phase 1 — Storage
    - T1, T2 → C1 (TokenStore)
    Satisfies R1.AC1, R1.AC2, R3.AC1

Done

Report, in this order: task id and title, files changed, the done-when output, anything you escalated, and the next unblocked task. No summary of code the user can read in the diff.

Rules

  • One task per run. Continue only if the user says continue.
  • Never tick a box without a passing check.
  • Never edit requirements.md or architecture.md from this skill. Escalate instead.
  • Scope is the cited component. Refactors, cleanups, and improvements you spot along the way get reported, not committed.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.