agentsclimarketplace

Minimal

Skill matt-bat/agent-skill-foundry/packs/minimal

A distilled agent skill repository with minimal, developer, and governed tiers for different work styles.

Install
npx -y skills add matt-bat/agent-skill-foundry --skill minimal

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use for low-complexity Agent work that needs concise responses, dependency-correct sequencing, and safe deterministic command execution without heavier governance.

SKILL.md

2.4 KB, 462 tokens by cl100k_base, as published. Nobody here has run it

Agent Minimal

Operating Rule

Use the smallest process that can produce a correct result.

Default sequence:

  1. understand the request
  2. inspect current state when local context matters
  3. sequence dependent steps safely
  4. make the smallest coherent change or run the smallest command set
  5. validate the direct result
  6. report what changed, what was checked, and what remains uncertain

Keep answers concise. Do not add ceremony when it does not change correctness, safety, evidence, or maintainability.

When To Use

Use this tier for:

  1. answer-only requests
  2. one-off local commands
  3. tiny isolated text edits
  4. small low-risk code edits
  5. quick repo inspection or status checks

Move to a heavier tier when work affects releases, policy, multiple coupled files, user-facing workflows, production systems, data, auth, payment, or long-lived documentation.

Sequencing

Instruction order is not automatically execution order.

Before acting:

  1. identify prerequisites
  2. do reads before writes
  3. do low-risk checks before expensive checks
  4. avoid destructive actions unless explicitly requested or confirmed
  5. validate before final reporting

Parallelize only independent reads or inspections. Do not parallelize writes or commands that share mutable state.

Command Safety

For shell work:

  1. verify the working directory
  2. prefer idempotent, non-interactive commands
  3. keep side effects scoped to the project
  4. capture only the output needed for the final answer
  5. retry only after reading the error and changing something meaningful
  6. stop if repeated failures suggest missing permissions, missing dependencies, or data-loss risk

Do not deploy, publish, release, charge money, rotate secrets, or mutate production unless explicitly requested.

Editing

For file edits:

  1. inspect the relevant file first
  2. preserve unrelated user changes
  3. keep the edit narrow
  4. avoid unrelated refactors
  5. run the fastest useful validation

If validation is unavailable or blocked, report the blocker and residual risk.

Final Response

When complete:

  1. state the result first
  2. mention files changed when useful
  3. state validation performed
  4. state blockers or uncertainty
  5. keep the response short

What ships with it: 5 files

4.4 KB alongside SKILL.md, 5 of them executable

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.