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 skills add ShaftHQ/SHAFT_ENGINE --skill architectgit clone --depth 1 https://github.com/ShaftHQ/SHAFT_ENGINEWrote 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.
[](https://agentmods.dev/skills/shafthq/shaft_engine/architect)<a href="https://agentmods.dev/skills/shafthq/shaft_engine/architect"><img src="https://agentmods.dev/badge/skills/shafthq/shaft_engine/architect/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.
<a href="https://agentmods.dev/skills/shafthq/shaft_engine/architect"><img src="https://agentmods.dev/badge/skills/shafthq/shaft_engine/architect.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00036 | $0.00980 |
| Opus 5.5 | $0.00014 | $0.00392 |
| Sonnet 5.5 | $0.00007 | $0.00196 |
| Haiku 4.5 | $0.00004 | $0.00098 |
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 — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Architect
Judgement cards for design-level work. Core triage sets depth; Ponytail still caps scope. Load ONE card file below, not all.
First: retrieve
python3 .chaos-engine/tool.py retrieve --store graphify "<module> callers dependencies": blast radius, seams, who breaks.python3 .chaos-engine/tool.py retrieve --store mempalace "<topic> decision ADR gotcha": prior decisions, rejected options, lessons.- Record
retrieve: used. Contradicts a prior decision: cite it, supersede it in the ADR, never silently diverge.
Tier → cards
| Triage tier | Use |
|---|---|
| One file, reversible | 18 debug, 19 review only |
| One module, reversible | add 1, 3, 9, 16, 17 |
| Public contract / hard to reverse | all relevant; 4 ADR mandatory |
Card index
| # | Card | Rule | File |
|---|---|---|---|
| 1 | Requirements | Testable acceptance criteria + non-goals before design. | analyze |
| 2 | Quality attributes | Name top 3 (ISO 25010) as measurable scenarios. | analyze |
| 3 | Trade-offs | 2+ options, criteria from 2, pick by reversibility. | analyze |
| 4 | ADR | Hard-to-reverse choice gets a 10-line ADR. | analyze |
| 5 | C4 sketch | Multi-component change: context + container diagram. | analyze |
| 6 | API contracts | Additive change; versioned; idempotent; Hyrum's law. | design |
| 7 | Boundaries | Domain inward, I/O outward; depend on stable things. | design |
| 8 | Domain language | Code names = domain words; one context, one model. | design |
| 9 | Principles | SOLID/DRY/KISS/YAGNI calibrated; rule of three. | design |
| 10 | Patterns | Pattern by forces present today, never speculative. | design |
| 11 | Failure modes | Timeout, bounded retry+jitter, idempotency, breaker. | operate |
| 12 | Observability | Structured logs, RED metrics, traces, SLO. | operate |
| 13 | Security | Trust boundaries, STRIDE-lite, least privilege. | operate |
| 14 | Performance | Measure first; budget; fix the bottleneck only. | operate |
| 15 | Data | Expand/contract migrations; one writer per fact. | operate |
| 16 | Test strategy | Risk-based mix; contract tests at boundaries. | change |
| 17 | Refactor/legacy | Characterize, then small safe steps; strangler. | change |
| 18 | Debugging | Root cause before fix; 3 failed fixes → rethink. | change |
| 19 | Review | Design > correctness > tests > clarity; severity. | change |
| 20 | Tech debt | Name it, price it, file it; pay where you touch. | change |
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 · 59 lines · 36 tokens per session scan A 5d65e2595dea
architect is a skill published in the GitHub repository ShaftHQ/SHAFT_ENGINE (416 stars, last pushed today), licensed MIT. It adds 36 tokens to every session and 980 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-10-18.
Other skills, from other repositories
UI Debug Agent
Skill inspect web/mobile applications bằng browser tools, phân tích DOM elements, xác định locators ổn định, debug UI automation failures, và hỗ trợ sinh Page Object classes.
Flaky Test Analyzer
Skill phân tích và khắc phục các automation test không ổn định (flaky tests), xác định root cause và đề xuất fix.
Locator Healer Agent
Skill tự động phát hiện và sửa chữa locators bị hỏng khi automation tests fail do thay đổi DOM.
nullaway
Guide for resolving NullAway static analysis errors. Best practices for: Passing ObservableSupplier/Supplier Dereferencing potentially @Nullable values Adding @NullMarked to Java code.
tika-eval-encoding-regression
Condensed tika-eval pattern for charset-detector regression hunts ("A picks encoding X, B picks Y") using one build and two configs — encoding-pair flip queries, OOV/languageness/FFFD signals, per-file detector attribution.
wxjava-troubleshooter
A troubleshooting guide for WxJava, a Java library used to connect applications to WeChat services. It organizes problems involving setup, access tokens, signatures, payment certificates, callbacks, data conversion, network requests, and multiple accounts.