Skip to content
Agentmart
Submit a tool

efficient-tool-use

Apply evidence-based rules when composing tool calls, parsing their output, interpreting non-zero exits, or turning repeated operations into durable routines.

Skill by emliunix

IndexedGPL-3.0
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/emliunix-efficient-tool-use

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Apply evidence-based rules when composing tool calls, parsing their output, interpreting non-zero exits, or turning repeated operations into durable routines. Use before or during tool-heavy work when a result will drive a decision, especially for transcripts, logs, structured output, guards, or service checks. Topics live under this one handle. Do not use for auditing a whole transcript corpus; use audit-agent-tooling for that procedure.

Use cases: security, agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found
Runs as
Node.js package

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/efficient-tool-use

What's inside

Skill name
efficient-tool-use
Description
Apply evidence-based rules when composing tool calls, parsing their output, interpreting non-zero exits, or turning repeated operations into durable routines. Use before or during tool-heavy work when a result will drive a decision, especially for transcripts, logs, structured output, guards, or service checks. Topics live under this one handle. Do not use for auditing a whole transcript corpus; use audit-agent-tooling for that procedure.
License
GPL-3.0
Instructions
39 lines in SKILL.md · view on GitHub

Before installing

  • Ships executable scripts

Files (7)

  • SKILL.md
  • call-composition.md · references/
  • output-shape.md · references/
  • result-interpretation.md · references/
  • reuse-and-protection.md · references/
  • capture-command-log.mjs · scripts/
  • check-capture-command-log.mjs · tests/

SKILL.md (first lines)

# Efficient tool use

One trigger, topic-grouped rules. Read only the topic that matches the moment;
the topic files are the durable rulebook, not this routing surface.

## Topics

- **`efficient-tool-use/output-shape`** -
  [output-shape.md](references/output-shape.md): ask the owning surface for
  structure, search before reading, parse records rather than words, and make
  required arguments explicit.
- **`efficient-tool-use/result-interpretation`** -
  [result-interpretation.md](references/result-interpretation.md): put the
  discriminating field in the test, bucket non-zero exits by code, keep crash
  rate separate from correctness, and split by command before blaming syntax.
- **`efficient-tool-use/call-composition`** -
  [call-composition.md](references/call-composition.md): keep payloads out of
  command strings, block on a condition instead of polling, and bound a command
  that spawns children by killing its process group rather than its parent.
- **`efficient-tool-use/reuse-and-protection`** -
  [reuse-and-protection.md](references/reuse-and-protection.md): save repeated
  operations, measure a guard's reach and activation, require identity rather
  than liveness, record the read time for live evidence, and capture raw output
  with its revision for hand-over instead of hand-typed transcripts.

Open the matching topic immediately when a result will be counted, parsed,
compared, or used as evidence. Do not rely on remembering a rule.

## Evidence

Source case: a six-transcript agent tool-use audit, 2026-10-02, based on
17,551 tool calls. The underlying report stays in its originating workspace;
this cookbook carries only the normalized rules and their bounds.

The source report and its parsers remain the first case study. Add a rule to a
topic only when it changes a future decision and its evidence is reproducible.
Replace superseded rules in place rather than keeping contradictory versions.

From emliunix/home.conf, GPL-3.0.

More from emliunix/home.conf

Signals

Stars
0
Forks
1
Open issues
3
Last commit (this folder)
Oct 14, 2026 (14 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License
GPL-3.0
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from emliunix

Cite this page

efficient-tool-use skill. Agent Mart. https://agentmart.dev/skills/emliunix-efficient-tool-use (last updated 2026-10-24)