arch

An architectural record describing how the Godpowers project is organised, including its skills, specialist agents, routing files, workflows, and state tracking.

In plain words
What is it for?
Use it when planning or reviewing changes to Godpowers' runtime architecture, command-line interface, routing, workflow plans, or project state handling.
Why use it?
It gives contributors a shared reference for important design decisions and where those decisions are represented in the codebase.

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/hannsxpeter/godpowers/arch
Clone the repo
git clone --depth 1 https://github.com/hannsxpeter/godpowers
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,479 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 $0.00000 $0.01479
Opus 5 $0.00000 $0.00740
Sonnet 5 $0.00000 $0.00296
Haiku 4.5 $0.00000 $0.00148

Measured 2d ago against content hash 7d9e34284b41, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

arch 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.

agents/arch.md · 88 lines

How it starts

The opening of the file, as written. The whole thing — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Scope

  • [DECISION] This pillar captures architectural context for Godpowers.

Context

  • [DECISION] Runtime architecture is implemented by source-controlled skills, routes, workflows, specialist contracts, schemas, and dependency-free Node.js helpers.

Decisions

  • [DECISION] Godpowers uses a pure-skill runtime where slash-command skills spawn specialist agents inside the host AI coding tool.
  • [DECISION] The CLI surface stays narrow: installer and uninstall flows, read-only status and next-route helpers, automation status and setup planning, dogfood fixtures, and extension scaffolding.
  • [DECISION] Routing decisions are stored in routing/*.yaml.
  • [DECISION] Workflow plans are stored in workflows/*.yaml and planned by lib/workflow-runner.js.
  • [DECISION] Workflow plans can expose visible local helpers such as repo-doc-sync, repo-surface-sync, host-capabilities, source-sync-back, and checkpoint-sync.
  • [DECISION] lib/state-lock.js provides cooperative advisory locking through .godpowers/state.json.
  • [DECISION] Existing .godpowers projects refresh runtime feature awareness through lib/feature-awareness.js.
  • [DECISION] ARCHITECTURE.md owns the architecture audit playbook for disconnected commands, actions, and workflows.
  • [DECISION] ARCHITECTURE-MAP.md renders the same audit as a graph from skills to routes, agents, workflows, recipes, docs, and package checks.
  • [DECISION] ARCHITECTURE-MAP.md keeps a complete core command supplement so all 124 shipped skills appear in the human-readable map.
  • [DECISION] Workflow plans use canonical helper IDs such as source-sync-back and pillars-sync-plan, while /god-sync output may show the shorter aliases source-sync and pillars-sync.
  • [DECISION] Godaudits 2.x interoperability reads .godaudits/AUDIT.json as canonical machine state, imports explicit check outcomes, evidence metadata, compliance, accepted risks, open questions, score caps, coverage, findings, and typed GA tasks, and uses generated or legacy AUDIT.mdx only as a fallback.
  • [DECISION] Godplans 1.1 interoperability treats .godplans/PLAN.mdx plus the pinned executable .godplans/validate-plan.sh as one contract, mirrors structural validation without executing repository shell during import, blocks GP dispatch outside approved or executing, and requires the official validator to pass immediately before work.
  • [DECISION] lib/prose-lint.js stays inside the existing artifact-quality boundary as a dependency-free inert-text scanner with no I/O, bounded sanitized excerpts, and warning-only U-12 integration through lib/have-nots-validator.js.
  • [DECISION] references/building/BLAST-RADIUS.md is the single Stage 2 safety-case protocol consumed by /god-review, /god-build, god-executor, god-quality-reviewer, and the orchestrator runbook.
  • [DECISION] god-executor proposes a candidate safety fact and focused proof but does not grade its own work; god-quality-reviewer independently verifies or replaces the candidate and owns the Stage 2 verdict.
  • [DECISION] lib/impact.js remains a static candidate generator with unproven maximum-level-2 metadata, while the existing evidence ledger remains the executed-proof authority.
  • [DECISION] lib/evidence.resolveReviewEvidence adds a read-only sanitized projection over unchanged verification records and additive record-ID and record-digest gate-event bindings.
  • [DECISION] Bounded changes receive 1 Stage 2 safety case; changes crossing at least 3 boundary classes or at least 2 high-impact classes receive at least 2 independent fresh-context safety cases before reconciliation.
  • [DECISION] The blast-radius design adds no command, route, recipe, workflow, specialist type, production dependency, evidence store, or state writer.
  • [DECISION] lib/verification-profile.js, lib/debug-feedback-loop.js, lib/program-design.js, lib/why-evidence.js, lib/events.js, and lib/invocation-policy.js add bounded evidence and authority contracts inside existing runtime boundaries.
  • [DECISION] Canonical route filenames own command identity, and route loading rejects duplicate, spoofed, or noncanonical YAML before invocation policy can be consumed.
  • [DECISION] @godpowers/operations-pack remains a lazy extension with no core route and no automatic external mutation authority.
  • [DECISION] The current executable audit status is fresh for repo surface, route quality, recipe coverage, and workflow planning.

Read the full file on GitHub · 88 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 · 88 lines · 0 tokens per session scan A 7d9e34284b41

Subscribe to this mod's changes

arch is an agent published in the GitHub repository hannsxpeter/godpowers (5 stars, last pushed 13d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,479 tokens. 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 2026-08-31.