goal-condition-synthesis

Community

Synthesize a falsifiable /goal from plan artifacts

Authorsnewhouse
Version1.0.0
Installs0

System Documentation

What problem does it solve?

It prevents LLM agents from drifting or looping by generating a testable Claude Code /goal completion condition grounded in observable plan artifacts rather than vague intent.

Core Features & Use Cases

  • Falsifiable goal construction: Produces a /goal condition that references specific transcript-observable signals (e.g., provenance.log lines, tasks.md status fields, git tags, and test exit codes).
  • Deterministic turn-cap budgeting: Computes or validates a capped number of turns based on pending milestones to bound cost for AFK execution or verification loops.
  • Canonical contracts for auditability: Owns the verdict-token vocabulary, observable-artifact list, and a deterministic condition-hash so other surfaces can link back without drift.

Quick Start

Use the goal-condition-synthesis Skill to synthesize a full-execute completion goal for an active AA-MA task when /execute-aa-ma-full reaches §2.5.

Dependency Matrix

Required Modules

None required

Components

Standard package

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: goal-condition-synthesis
Download link: https://github.com/snewhouse/aa-ma-forge/archive/main.zip#goal-condition-synthesis

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.