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 marconae/speq-skill --skill speq-writing-guardrailsgit clone --depth 1 https://github.com/marconae/speq-skillWrote 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/marconae/speq-skill/speq-writing-guardrails)<a href="https://agentmods.dev/skills/marconae/speq-skill/speq-writing-guardrails"><img src="https://agentmods.dev/badge/skills/marconae/speq-skill/speq-writing-guardrails/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/marconae/speq-skill/speq-writing-guardrails"><img src="https://agentmods.dev/badge/skills/marconae/speq-skill/speq-writing-guardrails.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00000 | $0.00966 |
| Opus 5 | $0.00000 | $0.00483 |
| Sonnet 5 | $0.00000 | $0.00193 |
| Haiku 4.5 | $0.00000 | $0.00097 |
Grade A, and why
speq-writing-guardrails 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 5d 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 — 53 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Writing Guardrails
Prose anchored in BLUF / Inverted Pyramid structure and informed by ASD-STE100 controlled-language discipline. Follow the following rules for governed prose.
Scope
- Govern: prose in
plan.md,spec.md,mission.md, decision logs, verification reports, GitHub PRs, issues, and comments. - Leave unchanged: Gherkin, Background bullets, tables, ASCII diagrams, delta markers, and validator-owned RFC keyword casing.
- PR text: lead with tradeoffs and boundaries; tie claims to diffs, files, tests, or command output.
Rules
- Classify each passage as procedural or descriptive.
- One sentence states one idea. Split a sentence joined by
and,which, orwhileinto two. - Procedural: imperative, one instruction per sentence, maximum 20 words.
- Descriptive: simple tense, one fact per sentence, one topic per paragraph, maximum 25 words and six sentences per paragraph.
- Put the conclusion first. Make headings summarize their sections. Make each section stand alone.
- State the current content, plan, or decision as fact. Do not narrate the steps, attempts, or revisions that produced it.
- Use active voice and name the actor. Use verbs, not nominalizations.
- Put conditions before commands:
If the build fails, read the log. - Use complete grammar. Keep articles and
that. Use no perfect or progressive forms. - Use no semicolons or contractions.
- Replace weak requirements:
should→MUSTor a fact. Replace hypothetical modals with conditions. Keep uncertainty when evidence requires it. - Use
MUST,SHOULD, andMAYonly as RFC 2119 / 8174 normative keywords. - Label destructive instructions:
WARNINGfor injury,CAUTIONfor data or equipment damage. State the consequence after the instruction. - Use one term, meaning, word class, and verb for each concept. Repeat nouns when pronouns lose the referent.
- Remove filler, vague claims, hedges, escape hatches, and
and/or. Quantify vague terms. - Ban hyperbole. Use no superlatives or intensifiers:
massive,game-changing,seamless,dramatically,very. State scale or impact only with a number or a test result. - Use one verb per concept:
check,run, andshow, or another chosen set. - Replace:
utilize→use;in order to→to;prior to→before;ensure→make sure that;facilitate→help;e.g.→for example;i.e.→that is;etc.→ named items;functionality→functionorfeature;out of the box→by default;under the hood→internally. - Keep technical names, product names, identifiers, commands, flags, paths, config keys, code, quoted errors, and numbers with units unchanged. Count each as one word.
- Do not claim project behavior without evidence. Scope unsupported claims, label inferences, and name untested areas.
- Ban em dashes. Use a comma, period, colon, parentheses, or a new sentence.
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.
- 5d ago Changed · -23 lines · -91 tokens per session 14a31837fd91
- 12d ago First seen · 76 lines · 91 tokens per session scan A f32d104f15ef
speq-writing-guardrails is a skill published in the GitHub repository marconae/speq-skill (50 stars, last pushed 6d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 966 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 2026-08-30.
Other skills, from other repositories
build-executor
Govern implementation from an approved execution contract. Invoke when execution-contract.md is approved and the user wants disciplined build work, TDD execution, or guarded batch-by-batch implementation.
workflow-start
Primary entry point for the spec-superflow state-machine workflow. Invoke when the user is inside an active spec-superflow change directory (look for .spec-superflow.yaml, changes/ /, proposal.md, specs/, design.md, tasks.md, or execution-contract.md) and asks to start, continue, resume, implement, plan, or figure out…
bug-investigator
Use when encountering any bug, test failure, or unexpected behavior during spec-superflow execution, before proposing fixes. Invoked automatically when build-executor hits a blockage.
release-archivist
Close out a spec-superflow change with verification, summary, and archive readiness. Invoke when implementation is complete, verification is underway, or the user asks for a final wrap-up.
contract-builder
Convert approved planning artifacts into an execution contract. Invoke when the user wants to start building, asks to move from planning to implementation, or when execution-contract.md is missing or stale.
need-explorer
Clarify intent, scope, constraints, and success criteria before artifact creation. Invoke when the request is fuzzy, the user is comparing options, or the workflow needs a stable change definition before writing artifacts.