CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

bmad-build

Implements a story, feature, or bug fix end to end: plan, code, review, verify. Use when the user hands one over and leaves the edits to you; a bare ticket id or link counts. Skip obvious, low-risk mechanical maintenance such as small ignore-file, typo-only, formatting-only, or configuration-hygiene edits. Explicit BMAD requests always qualify. Do not volunteer for user-directed interactive edits or version-control operations that only record existing work.

Quality

93%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

96%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

An exemplary thin-dispatcher skill: fully executable command with precise substitution rules, explicit HALT and failure gates, and a recovery path, all in a lean body. The only gap is that the bundled reference files are not mentioned or linked in the body, leaving their navigation implicit in the render step.

DimensionReasoningScore

Conciseness

The ~15-line body is a lean dispatcher: one command, two substitution rules, flag mappings, and failure handling, with zero explanation of concepts Claude already knows — every clause ("exactly once", "without changing the current working directory") carries a guard, matching the 5 anchor where every token earns its place.

5 / 5

Actionability

The command is fully executable with both placeholders given unambiguous resolution rules ("nearest folder containing `_bmad/`, from the working directory upward"; "absolute path to this skill's directory"), and every branch — route flag, review flag, HALT, missing script, uv unavailable — has exact instructions. Not 4: there are no gaps in the guidance; all common invocation cases are covered.

5 / 5

Workflow Clarity

The sequence is unambiguous with explicit checkpoints: run once → apply invocation-named flags → follow printed workflow, "HALT: <reason>: report the reason and stop" as a hard validation gate, and a recovery loop for the missing-script case (offer setup, "then run the command again") plus a stop on any other failure. This is a simple under-50-line single-purpose skill whose every branch is specified, satisfying the simple-skill exception at full marks.

5 / 5

Progressive Disclosure

The bundle structure is healthy — a thin SKILL.md dispatcher plus two one-level-deep reference files (references/claims-check.md, references/deletion-check.md) with nothing inlined that belongs elsewhere — but the body never signals those reference files; their discovery depends entirely on the external render step, a minor organization gap. Not 5: references are not clearly surfaced in the SKILL.md itself; not 3: nothing that should be separate is inline and the structure is clean.

4 / 5

Total

19

/

20

Passed

Description

87%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: third-person, concise, with an explicit trigger clause and unusually disciplined include/exclude boundaries. The only softness is that the capability verbs are workflow-stage abstractions and a few natural synonyms (issue, task) are absent.

DimensionReasoningScore

Specificity

"Implements a story, feature, or bug fix end to end: plan, code, review, verify" names the domain and lists four concrete action stages, but the verbs (plan, code, review, verify) stay at the workflow-stage level rather than the fully concrete operations of the 5 anchor, and coverage of what those stages entail is not spelled out.

4 / 5

Completeness

Both halves are explicit: the what ("Implements a story, feature, or bug fix end to end: plan, code, review, verify") and the when ("Use when the user hands one over and leaves the edits to you; a bare ticket id or link counts"), reinforced with explicit exclusions ("Skip obvious, low-risk mechanical maintenance...", "Do not volunteer for..."). This matches the 5 anchor's concrete-trigger-phrasing standard; not 4 because the when-clause is maximally explicit, not merely adequate.

5 / 5

Trigger Term Quality

Natural trigger terms are present — "story", "feature", "bug fix", "a bare ticket id or link counts", "Explicit BMAD requests" — which users would plausibly say, but common synonyms like "issue", "task", or tracker names are missing, so coverage is good rather than comprehensive.

4 / 5

Distinctiveness Conflict Risk

The BMAD niche is distinct and the description actively draws boundaries in both directions ("Explicit BMAD requests always qualify", skip maintenance edits, do not volunteer for interactive edits or version-control operations), minimizing the chance of triggering for the wrong skill. Not 4: the overlap risk with generic implementation requests is explicitly fenced off rather than left minor.

5 / 5

Total

18

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

bmad-code-org/BMAD-METHOD

Path

skills/bmad-build/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.