Scaffold, interview, audit, reconcile, or seed Ideal State Artifacts with verification evidence.
npx skills add https://github.com/davdunc/pai-framework --skill isa-davdunc
Please help me install this Agent Skill. Skill: ISA Source: https://github.com/davdunc/pai-framework/tree/main/skills/ISA Command: npx skills add https://github.com/davdunc/pai-framework --skill isa-davdunc
The ISA skill solves the problem of vague or changing expectations by turning a goal or requirement into a hard-to-vary Ideal State Artifact that can be built, tested, reconciled, and verified across iterations.
Request: "Run ISA Scaffold to create an Ideal State Artifact for my project at tier E3, using this prompt: [paste your prompt]."
High-intent search queries and answers about installing and using this skill.
To turn ambiguous requirements into verifiable specifications, scaffold an Ideal State Artifact that defines done as testable criteria rather than opinions. This locks section order and applies tier completeness gates to keep expectations stable across iterations.
An Ideal State Artifact is a canonical specification document that defines done as verifiable criteria. You need one for non-trivial feature building, requirement hardening, and task decomposition where completion must be testable and stable across iterations.
To merge ephemeral feature slices back into a master specification deterministically, use a reconciliation workflow that leverages stable ISC IDs. This ensures updates are merged accurately without overwriting the locked canonical structure.
Yes, you can bootstrap a project specification from an existing repository using a seeding workflow. This process generates an Ideal State Artifact by extracting existing structure and applying verification evidence requirements to your current codebase.
You audit specification completeness against tier gates by running a dedicated checking workflow that scores the document. This evaluates whether the artifact meets explicit tier completeness gates and identifies missing sections requiring further deepening.
The best way to fill gaps in a technical specification is through an adaptive Q&A interview process. This deepens missing sections by asking targeted questions to ensure the final artifact meets strict completeness gates and verification requirements.
alivecontextScaffold ALIVE walnut folders with _kernel files from SKILL.md frontmatter.
michael-conradValidate success criteria and collect artifacts before task completion.
BellaBeEvaluate artifacts against criteria and output verdicts with evidence tables.
dimagi-internalRun eight deterministic QA checks on solicitation review artifacts.
