tooltest-fix-loop

Use when running tooltest to validate MCP servers, interpret failures, and iterate fixes in this repo.

10 stars

Best use case

tooltest-fix-loop is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

Use when running tooltest to validate MCP servers, interpret failures, and iterate fixes in this repo.

Teams using tooltest-fix-loop should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

  • You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

  • You only need a quick one-off answer and do not need a reusable workflow.
  • You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/tooltest-fix-loop/SKILL.md --create-dirs "https://raw.githubusercontent.com/lambdamechanic/tooltest/main/skills/tooltest-fix-loop/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/tooltest-fix-loop/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How tooltest-fix-loop Compares

Feature / Agenttooltest-fix-loopStandard Approach
Platform SupportNot specifiedLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

Use when running tooltest to validate MCP servers, interpret failures, and iterate fixes in this repo.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

SKILL.md Source

# Tooltest fix loop

You have access to this repository and can run commands.

Goal: make the repository's MCP server(s) conform to the MCP spec as exercised by tooltest.

Figure out how to start the MCP server from this repo (stdio or streamable HTTP).

Select a small, related subset of tools intended to be used together. Default to testing at most 50 tools at a time, and strongly prefer a smaller group. Use `--tool-allowlist` (or `tool_allowlist` in MCP input) to enforce this.

Run tooltest against it and fix failures until it exits 0.

If you see "state-machine generator failed to reach minimum sequence length", re-run with `--lenient-sourcing` or seed values in `--state-machine-config`.

CLI usage (preferred when you can run commands):
- Use CLI-only flags for debugging, e.g. `--trace-all /tmp/tooltest-traces.jsonl`.
- Examples:
  CLI stdio (allowlist example): `tooltest stdio --command "<command that starts the repo's MCP server>" --tool-allowlist foo --tool-allowlist bar`
  CLI http (allowlist example): `tooltest http --url "http://127.0.0.1:9000/mcp" --tool-allowlist foo --tool-allowlist bar`

MCP tool usage (when you must call via MCP):
- Call the `tooltest` tool with the shared input schema.
- Only fields in the MCP input schema are accepted (CLI-only flags like `--json` and `--trace-all` are not supported).
- Example (allowlist):

```json
{
  "target": { "stdio": { "command": "<command that starts the repo's MCP server>" } },
  "tool_allowlist": ["foo", "bar"]
}
```

Don't rename tools or change schemas unless required; prefer backward-compatible fixes.

Add/adjust tests if needed.

Return a short summary of what you changed and why, plus the final passing tooltest output snippet.

Related Skills

testing-patterns

10
from lambdamechanic/tooltest

Testing patterns and standards for this codebase, including async effects, fakes vs mocks, and property-based testing.

rust-guidelines

10
from lambdamechanic/tooltest

Pragmatic Rust conventions to keep code readable, testable, and performant for this project.

lambda-workflow

10
from lambdamechanic/tooltest

One lifecycle for Lambda repos: choose a br issue, start work, land the PR, and watch GitHub via Dumbwaiter MCP until it merges.

galahad

10
from lambdamechanic/tooltest

how to approach tests, types and coverage

beads

10
from lambdamechanic/tooltest

Git-backed issue tracker for multi-session work with dependencies and persistent memory across conversation compaction. Use when work spans sessions, has blockers, or needs context recovery after compaction.

bd-to-br-migration

10
from lambdamechanic/tooltest

Migrate docs from bd (beads) to br (beads_rust). Use when updating AGENTS.md, converting bd commands, "bd sync" → "br sync --flush-only", or beads migration.

catcolab-causal-loop

16
from plurigrid/asi

CatColab Causal Loop Diagrams - systems dynamics modeling with reinforcing (R) and balancing (B) feedback loops, delays, and Lotka-Volterra semantics for strategic analysis.

self-validation-loop

16
from plurigrid/asi

Run self-validation loops for triadic color systems using prediction vs observation and error minimization.

repo-change-loop

16
from woojubb/robota

Runs the standard Robota change loop by identifying impact, building affected scope, running targeted verification, and summarizing residual risk. Use when making or reviewing repository changes that should end in an explicit verification result.

cfn-loop-validation

14
from masharratt/claude-flow-novice

Multi-layer validation and quality gates for CFN Loop workflows. Use when implementing gate checks, consensus validation, or enforcing clean agent exit patterns.

cfn-loop-output-processing

14
from masharratt/claude-flow-novice

Type-safe output processing for Loop 2 validators and Loop 3 implementers. Use when parsing agent confidence scores, feedback, or calculating consensus from multiple validators.

cfn-loop-orchestration-v2

14
from masharratt/claude-flow-novice

CFN Loop coordination and orchestration - gate checks, validation, consensus, Product Owner decisions. Use when orchestrating multi-agent workflows, managing iteration cycles, or coordinating Loop 2/Loop 3 dependencies.