Writing Tech Post

What the source does

• FilesystemNone observedNo reads or writes in the source
• NetworkNone observedNo outbound calls in the source
• ShellNone observedNo commands run in the source
• DependenciesPinnedNothing floats
• SecretsNone foundNo keys in source

Five audit layers

Gates applied: no behavioural pass.

*The diagram below outlines the high-level architecture of [SYSTEM]. Anything outside the dashed box is out of scope for this post.*

*Figure N: High-level overview of [SYSTEM] node. [Component A] receives [DATA] from [SOURCE]; [Component B] forwards [PROCESSED DATA] to [DESTINATION]; [Component C] persists [STATE] in [STORE].*

# [AI/AGENT TITLE — "[SYSTEM-NAME]: A [CAPABILITY] agent" or "How we built [SYSTEM]"]

[LLM-generated [OUTPUT] carries the risk of [FAILURE MODE].] [In our evaluation, [SYSTEM] would sometimes [INCORRECT BEHAVIOUR].]

Audit 0.4.1 · source sha 118e61cf1c61 · 16 Oct 2026Report an issue or request a re-scan

• Publisherpedronauck GitHub userOwner of the repository this skill is read from
• Source folderpedronauck/skills/skills/mine/writing-tech-postThe exact folder we audited
• LicenceNot declaredAs the repository states it
• Commit pinned for this verdict118e61cf1c61A new audit runs when the source changes
• Repository statusNot archivedLast push 31 days ago
• Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

Its description: Write or edit engineering posts with supported claims. Use for drafts, rewrites, and publication reviews; excludes marketing pages, API references, release notes, and internal specs.

---
name: writing-tech-post
description: "Write or edit engineering posts with supported claims. Use for drafts, rewrites, and publication reviews; excludes marketing pages, API references, release notes, and internal specs."
metadata:
  author: Pedro Nauck
  github: https://github.com/pedronauck
  repository: https://github.com/pedronauck/skills
---
# Writing Engineering Posts

Produce the requested artifact: a new post, an outline, a focused rewrite, or a publication review. Reuse the accepted brief and evidence. Choose a structure that fits the reader and material; a small edit does not restart an entire authoring workflow.

1. Establish the reader, main point, and supported facts from available context. Ask only for missing information that materially affects the requested output.
2. For a new post, choose a useful primary archetype and outline. For an existing draft, retain working structure and edit the requested scope. Frontmatter, rung labels, depth tuples, and per-section evidence plans are optional tools unless the publisher requires them.
3. Explain the mechanism and why it matters. Support quantitative, comparative, security, and capability claims with appropriate evidence and limitations. Distinguish actual measurements, illustrative examples, hypotheses, and work still in progress.
4. Refine voice and flow where needed. Publisher style matrices, H2 question chains, and first/last-200-word comparisons are editorial heuristics, not universal gates. Do not invent failed mitigations, partial victories, benchmark results, or lessons to satisfy an arc.
5. Before external publication, apply `references/pre-publish-checklist.md` to the relevant claims and disclosure obligations. Draft delivery is not publication authorization. The optional linter reports heuristics; human factual review owns unsupported claims and disclosure decisions.

## Reference router

Read the reference for the current archetype or editing concern, not every phase:

- Structure: `references/archetypes-and-structure.md`; optional depth diagnosis: `references/depth-and-abstraction.md`.
- Incident: `references/postmortems.md`; migration: `references/migrations.md`; performance: `references/performance-deep-dive.md`.
- AI/agents: `references/ai-and-agents.md`; security/reliability: `references/security-and-reliability.md`.
- Evidence/assets: `references/evidence-diagrams-code.md`; narrative: `references/narrative-and-pacing.md`.
- Voice/disclosure: `references/voice-and-disclosure.md`; a requested publisher register: `references/publisher-voice-matrix.md`; specific prose problems: `references/anti-patterns.md`.

Use an `assets/outline.<archetype>.md` template only when useful; omit irrelevant sections. Run helpers through this skill's absolute directory: `python3 <writing-tech-post-dir>/scripts/lint-post.py <draft.md>`. `--strict` makes heuristic warnings fail for an explicitly chosen editorial gate. A clean lint result does not prove factuality or disclosure readiness.

Excerpt, first 2,998 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

git clone https://github.com/pedronauck/skills
git -C skills checkout 118e61cf1c61

Works in

1. 2026-10-16
   ✕BLOCK D 69 / 100

   first audit · source 118e61cf1c61

Questions about Writing Tech Post

• github.com/pedronauck/skills/tree/main/skills/mine/writing-tech- — the audited source at commit 118e61cf1c61 and every finding aboveObserved · 16 Oct 2026

Spotted something wrong? Report a correction · Methodology