spec-converge

spec-converge is a skill for Claude Code, Codex from JKHeadley/instar. It costs 135 tokens per session (10,921 once invoked), scanned A, original, MIT.

An iterative review process for a software-development specification. It asks several reviewers to examine security, growth, integration, unusual failure cases, and missing decisions before coding begins.

In plain words
What is it for?
Use it before substantial development work to audit a specification, address findings, and produce a final specification with a review report.
Why use it?
It helps find architectural and implementation problems that a single review may miss, then revises the specification until the reviews converge.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: mentions subagents; mentions Codex; mentions Gemini CLI.

Good fit Use it before substantial development work to audit a specification, address findings, and produce a final specification with a review report.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/jkheadley/instar/spec-converge
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add JKHeadley/instar --skill spec-converge
Clone the repo
git clone --depth 1 https://github.com/JKHeadley/instar

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for spec-converge

README.md
[![agentmods](https://agentmods.dev/badge/skills/jkheadley/instar/spec-converge/github.svg)](https://agentmods.dev/skills/jkheadley/instar/spec-converge)
Your own site
<a href="https://agentmods.dev/skills/jkheadley/instar/spec-converge"><img src="https://agentmods.dev/badge/skills/jkheadley/instar/spec-converge/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for spec-converge

Your own site · 80×15
<a href="https://agentmods.dev/skills/jkheadley/instar/spec-converge"><img src="https://agentmods.dev/badge/skills/jkheadley/instar/spec-converge.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 135 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 10,921 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector warn 7 Sept 2026
SkillSpector: 4 findings, up to high

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

  • high YARA Match · line 94
    YARA rule matched a known malware signature (reverse shell, backdoor, ransomware, C2 framework, or info stealer).
    Fix: Remove the malware payload or compromised file entirely. Investigate how it entered the skill and audit all other artifacts for additional indicators of compromise.
  • high Memory Poisoning · line 96
    Skill manipulates agent memory, state, or stored context. Memory corruption can alter personality, override safety rules, or cause unpredictable behavior.
    Fix: Protect agent memory and state from modification by untrusted content. Use read-only memory for critical instructions and validate all state changes.
  • medium Data Exfiltration · line 71
    Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
    Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
  • medium Excessive Agency · line 340
    Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
    Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00135 $0.10921
Opus 5 $0.00068 $0.05461
Sonnet 5 $0.00027 $0.02184
Haiku 4.5 $0.00014 $0.01092

Measured 9d ago against content hash e0e1129ade0f, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

spec-converge scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.

The scan reads SKILL.md. This mod also ships 3 executable files (scripts/cross-model-review.mjs, scripts/publish-spec-review.mjs, scripts/write-convergence-tag.mjs), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -sS -m 90 -X POST -H "Authorization: Bearer ${AUTH:-$INSTAR_AUTH_TOKEN}" -H 'Content-Type: application/json' \
skills/spec-converge/SKILL.md · 364 lines

How it starts

The opening of the file, as written. The whole thing — 364 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/spec-converge

Audience: the instar-developing agent. End users do not invoke this. Throughout this document, "the agent" refers to the instar-developing agent running the skill.


What this skill does

Takes a spec file, runs multiple parallel reviewers against it, updates the spec to address their findings, runs another round, and repeats until convergence. Produces a final converged spec and a comprehensive human-readable report.

The purpose is to catch architectural, security, adversarial, scalability, and integration issues BEFORE code is written, not after. The single-reviewer pattern used by /instar-dev's existing second-pass step has proven too narrow in practice — a four-reviewer parallel audit on the integrated-being ledger PR surfaced 14 serious issues that a single reviewer had missed. This skill bakes that multi-angle pattern in as the structural default for any substantive instar-dev work.

When it runs

Before /instar-dev touches any instar source, the change's spec must pass through this skill. /instar-dev's pre-commit hook refuses work unless:

  1. The spec file has a review-convergence: <timestamp> entry in its frontmatter, written by this skill on successful convergence.
  2. The spec file has an approved: true entry in its frontmatter, written by the user after reading the convergence report.

Both tags must be present. Without them, /instar-dev is blocked.

Input

One argument: the path to the spec file, relative to the instar repo root.

The spec file is a markdown document with YAML frontmatter. Minimum required structure:

---
title: "Short title"
slug: "url-friendly-slug"
author: "agent name"
---

# Title

## Problem statement
[what is being built and why]

## Proposed design
[how it works, at enough detail to review]

## Decision points touched
[any block/allow/route gates the design introduces, removes, or modifies]

## Open questions
[things that need human input]

The skill adds review-convergence, review-iterations, review-completed-at, and (if provided) approved fields to the frontmatter on successful convergence.

Read the full file on GitHub · 364 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 9d ago First seen · 364 lines · 135 tokens per session scan A e0e1129ade0f

Subscribe to this mod's changes

spec-converge is a skill published in the GitHub repository JKHeadley/instar (79 stars, last pushed today), licensed MIT. It adds 135 tokens to every session and 10,921 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

javascript-sast

JavaScript and Node.js security scanning. Checks dependency vulnerabilities via npm audit and source patterns for XSS, eval, and prototype pollution.

vladkesler/initrunner · 31 tokens

sdk

Guide users building apps, scripts, CI pipelines, or automations on top of the Cursor TypeScript SDK (@cursor/sdk). Use when the user mentions integrating, installing, or writing code against the Cursor SDK; says Agent.create, Agent.prompt, Agent.resume, agent.send, run.stream, CursorAgentError, or @cursor/sdk; asks…

Kevin-Liu-01/Agent-Machines · 178 tokens

jest-patterns

Jest and Vitest testing patterns including describe/it blocks, expect matchers, mocking, and async test strategies for JavaScript and TypeScript.

vladkesler/initrunner · 34 tokens

debugging-typescript-errors

Systematically investigates TypeScript compiler errors, traces the real type mismatch, and applies a minimal verified fix. Use when the user wants to debug TypeScript type errors.

dork-labs/dorkos · 39 tokens

code-review

Use when reviewing a code change or diff for correctness, security, missing tests, and convention violations before opening or approving a PR. Review independently and adversarially, then fix high-confidence issues.

SebaBoler/vanguard · 42 tokens

tech-spec

Use when a task is under-specified and needs a written technical specification BEFORE any code is written or changed. This skill only researches and documents — it never edits source files. Do not invoke for implementing, reviewing, or simplifying existing code.

SebaBoler/vanguard · 51 tokens