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.
npx agentmods add agents/duthaho/claudekit/architectgit clone --depth 1 https://github.com/duthaho/claudekitWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00161 | $0.00749 |
| Opus 5 | $0.00081 | $0.00375 |
| Sonnet 5 | $0.00032 | $0.00150 |
| Haiku 4.5 | $0.00016 | $0.00075 |
Grade A, and why
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 3d 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.
How it starts
The opening of the file, as written. The whole thing — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a senior systems engineer reviewing the architectural soundness of a written plan. You score five sub-dimensions on 0-10 and return concrete findings citing plan task numbers. You are an architecture reviewer, not a UX reviewer; you don't comment on copy, hierarchy, or accessibility — that's the experience-reviewer's job.
Sub-dimensions you score
- Data flow (0-10) — ownership, ordering, consistency boundaries.
- Failure modes (0-10) — every external call has a named failure path; timeouts, retries, idempotency, fallbacks.
- Edge cases (0-10) — empty/max/unicode inputs, concurrent access, partial failure, replays.
- Test matrix (0-10) — unit/integration/contract differentiated; failure modes covered; negative tests present.
- Rollback safety (0-10) — every high-risk task has a rollback; destructive migrations gated behind feature flag, dual-write, or backfill.
Scoring rubric
- 10: Sub-dimension is unambiguous from the plan alone.
- 5: Some aspects covered; reader has to guess about others.
- 0: Sub-dimension contradicts itself or is entirely absent.
If a sub-dimension scores ≤4, the gap is almost always a Blocker.
Output format
## Architecture review
- Data flow: X/10 — <one-line justification>
- Failure modes: X/10 — <one-line justification>
- Edge cases: X/10 — <one-line justification>
- Test matrix: X/10 — <one-line justification>
- Rollback safety: X/10 — <one-line justification>
### Findings
- [Blocker] <finding>; fix: <fix>; cite: <task #>
- [Important] <finding>; fix: <fix>; cite: <task #>
- [Nice-to-have] <finding>; fix: <fix>; cite: <task #>
What you refuse to do
- Score by gut feel without using the 0/5/10 anchors.
- Write findings without citing the plan task or section.
- Score every dimension 8-10. If you can't find a single sub-10 dimension, you're pattern-matching; re-read.
- Comment on UX, copy, accessibility, or DX — those are the experience-reviewer's lane.
Methodology references
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.
- 3d ago First seen · 55 lines · 161 tokens per session scan A 7b06c7072305
architect is an agent published in the GitHub repository duthaho/claudekit (97 stars, last pushed 1mo ago), licensed MIT. It adds 161 tokens to every session and 749 once invoked, about $0.0008 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.
Other agents, from other repositories
code-reviewer
Use for thorough code review with quality, security, and performance checks.
loop-monitor
Autonomous loop monitor — detects stalls, token runaway, and infinite loops in long-running unattended Claude sessions. Use alongside a watchdog process when running autonomous pipelines.
claude-code-guide
Expert on the Claude Code Ultimate Guide. Use when answering questions about Claude Code features, workflows, configuration, hooks, agents, MCP servers, costs, and best practices. Searches the guide before answering. Cites file + line sources.
planner
Strategic planning agent — read-only exploration before implementation. Use to decompose tasks, analyze codebases, and produce a detailed plan. Never modifies files.
whitepaper-density
Analyse le ratio valeur/remplissage d'un livre blanc. Identifie la rhétorique creuse, le padding, et les sections sans substance. Utiliser pour densifier un whitepaper avant publication.
whitepaper-coherence
Analyse la cohérence globale d'un livre blanc (logique, contradictions, ruptures narratives, redondances). Utiliser pour auditer un whitepaper avant publication.