Skip to content
SkillModel-led816 tokens in use

Spec Miner

Reverse-engineers a specification from an undocumented codebase: explores it read-only, traces data flows and writes EARS requirements.

12,471 stars on the repository MIT LicenseJeffallan/claude-skills · skills/spec-miner
/plugin install fullstack-dev-skills@fullstack-dev-skills
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing Spec Miner costs

101tokens idle
816tokens in use
2,817tokens, up to

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Up to adds the files its instructions link. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

Nothing detected: no shell, network, file writes, git or credentials in its files.

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Excellent
  • Frontmatter15/15
  • Structure22/25
  • Examples9/15
  • Output10/15
  • Size15/15
  • Resources15/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent

No findings.

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Excellent
  • Development activity100
  • Maintenance100
  • Community60
  • Maturity36
  • Momentum40

Spec Miner helps the agent make sense of legacy or inherited code with no documentation. It scopes the analysis, explores with Glob, Grep and Read, traces data flows and request paths, and writes observed requirements in EARS format with uncertainties flagged.

The skill works from two perspectives, an architecture view for structure and data flow and a QA view for observable behavior and edge cases, and it requires every observation to point to code evidence.

What You Get

  • A five-step reverse-engineering workflow with an exploration checkpoint
  • Example Glob and Grep patterns for entry points, TODO markers, configuration use and API routes
  • A quick reference for EARS requirement types with examples
  • A specification file saved as specs/{project_name}_reverse_spec.md
  • A seven-part output: stack, structure, requirements, non-functional notes, acceptance criteria, uncertainties and recommendations
  • Reference files for the analysis process, EARS format, specification template and analysis checklist

Common Use Cases

  • Onboarding to an inherited project with no documentation
  • Documenting an API from its route definitions
  • Finding technical debt markers and undocumented business rules
  • Preparing a legacy system for an enhancement or rewrite

What Makes It Unique

It restricts the agent to Read, Grep and Glob in its frontmatter, so analysis cannot change the code. It separates observed facts from inferences, requires a code location for each observation and gives uncertainties their own section, which suits a task where guessing is the main risk.

Under The Hood

The skill is a SKILL.md with four reference files that the agent loads by topic. It ships no scripts. Its frontmatter allows only the Read, Grep and Glob tools, so exploration is read-only. The one write is the final specification, which the instructions say to save under a specs folder in your project, and that file write needs the agent’s file-writing tools to be available.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers