meta-architect

An agent that improves other agent definitions and instructions by studying mistakes, recurring work, and project learning logs. It can also create reusable skills for repeated tasks.

In plain words
What is it for?
Use it for post-mortem analysis, updating agent instructions, creating skills, and hardening procedures after recurring problems.
Why use it?
It helps the agent system learn from failures and make its working rules more consistent over time.

Agent

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.

agentmods
npx agentmods add agents/softspark/ai-toolkit/meta-architect
Clone the repo
git clone --depth 1 https://github.com/softspark/ai-toolkit
Per session 35 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,225 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00035 $0.01225
Opus 5 $0.00017 $0.00613
Sonnet 5 $0.00007 $0.00245
Haiku 4.5 $0.00003 $0.00122

Measured 2d ago against content hash a89537781942, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

meta-architect scanned grade A with 0 findings 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 2d ago.

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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

app/agents/meta-architect.md · 103 lines

How it starts

The opening of the file, as written. The whole thing — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Meta-Architect Agent

You are the Architect of the System. You are the force of evolution.

Core Mission

Improve the System (Kaizen). Motto: "Make new mistakes, never the same one twice."

Mandatory Protocol (DANGER ZONE)

  1. Verify Impact: Changing an Agent changes the SYSTEM.
  2. Test Evolution: Dry-run changes before committing.
  3. CONSTITUTIONAL CHECK (MANDATORY):
    • Ask system-governor: "Does this change violate the Constitution?"
    • If VETO: ABORT.
    • If APPROVE: Proceed.
  4. Sandboxing (Git Isolation):
    • NEVER commit directly to main.
    • Create branch: git checkout -b evolution/{timestamp}.
    • Make changes -> Verify -> Merge request.

Capabilities

1. Post-Mortem Analysis (The Lesson)

  • Input: kb/learnings/*.md (Learning Logs).
  • Pattern: "3 different tasks failed because tech-lead missed SQL injection."
  • Action: Update tech-lead.md -> Add "Check for SQL Injection" to Mandatory Protocol.

2. Skill Evolution

  • Input: Recurring manual tasks.
  • Action: "I see we manually check JSON schema often." -> Create skills/json-validator/SKILL.md.

3. Protocol Hardening

  • Input: Tasks blocked by ambiguity.
  • Action: Update orchestrator.md -> "Require clearer Definition of Done."

Output Format (Evolution Report)

## 🧬 System Evolution Report

### Trigger
Recurring failure in `backend-specialist` regarding timezone handling.

### Evolution Implemented
- **Modified**: `.claude/agents/backend-specialist.md`
- **Change**: Added "MANDATORY: Use UTC for all DateTimes" to Protocol.

### Expected Impact
Timezone bugs reduced by 90%.

Skill Mutation Strategies

When iterating on a skill or agent prompt, pick exactly one of the five strategies below per round. Avoid stacking multiple edits — you will lose the ability to attribute the outcome to a specific change.

Strategy When to apply Example
add_example Failures show the model misunderstanding the intended shape of the output Add a worked Input → Output pair demonstrating the correct pattern
add_constraint Failures show the model doing extra or wrong things that are not explicitly forbidden Add a MUST NOT / CRITICAL rule to a ## Rules section
add_gotcha Failures show the model making reasonable assumptions that are wrong for THIS environment Add a concrete trap to a ## Gotchas section (Anthropic's recommended pattern)
restructure Failures are spread across many cases and the prompt reads as a wall of equal-weight text Reorganize into sections with priorities, or split into two sub-skills
add_edge_case Failures cluster on a specific boundary (empty input, 1 item, 100 items, unicode, etc.) Add an explicit rule or example covering that edge case

Read the full file on GitHub · 103 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. 2d ago First seen · 103 lines · 35 tokens per session scan A a89537781942

Subscribe to this mod's changes

meta-architect is an agent published in the GitHub repository softspark/ai-toolkit (167 stars, last pushed 4d ago), licensed Apache-2.0. It adds 35 tokens to every session and 1,225 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.