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 davidomil/cloudx --skill source-grounded-writinggit clone --depth 1 https://github.com/davidomil/cloudxWrote 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/davidomil/cloudx/source-grounded-writing)<a href="https://agentmods.dev/skills/davidomil/cloudx/source-grounded-writing"><img src="https://agentmods.dev/badge/skills/davidomil/cloudx/source-grounded-writing/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/davidomil/cloudx/source-grounded-writing"><img src="https://agentmods.dev/badge/skills/davidomil/cloudx/source-grounded-writing.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 | $0.00030 | $0.02256 |
| Opus 5.5 | $0.00012 | $0.00902 |
| Sonnet 5.5 | $0.00006 | $0.00451 |
| Haiku 4.5 | $0.00003 | $0.00226 |
Grade A, and why
source-grounded-writing 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 4d 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 — 128 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Source Grounded Writing
Use this skill for technical docs, design decisions, issues, MR comments, email, chat replies, and rendered artifacts. Match the work to the requested output. Evidence quality does not depend on document length.
Choose the writing path
Direct writing: For a short reply, issue, email, MR comment, or wording-only revision, read the relevant sources, check the claims, and write the requested text directly. Do not create IR files, probe rendering dependencies, compile Quarto, or install tools. A one-sentence answer may still need substantial investigation.
Artifact writing: Use the IR and compiler for substantial, reusable, or rendered documents when the evidence structure and formats justify them. Render only requested targets. Read setup and migration instructions in README.md when using this path.
Both paths follow the same source policy and meaning checks. Follow the session's documentation-search and original-source ingestion rules; this skill does not replace them. If archive services are unavailable, report that and inspect available originals without claiming archive validation.
Establish the brief
Identify these from the request and prior conversation before drafting:
- What does the reader know, and what must they decide or do?
- What output, length, scope, and task order did the user request?
- What has the user excluded or asked to remove?
- Which source revision, configuration, board, or run is relevant?
Infer routine choices from context. Ask only when missing information materially changes the result. Carry corrections forward for the same artifact or channel: if testing narration was removed, do not add it back in a later rewrite. Keep qualifications when removing them would change meaning. Separate new requirements and implementation changes from avoidable editorial rework.
For a beginner guide, explain prerequisites in task order and try each instruction from that reader's perspective. Introduce terms before using them. Keep project history and optional future features outside the operator's main path.
What ships with it
26 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.
- .gitignore 67 B
- assets/fixtures/minimal-ir.json 4.7 KB
- assets/fixtures/reset-design-note.txt 340 B
- assets/quarto/_quarto.yml 174 B
- assets/quarto/source-grounded.css 434 B
- package-lock.json 255 KB
- package.json 880 B
- README.md 9.2 KB
- references/evidence-policy.md 4.0 KB
- references/intermediate-schema.json 16 KB
- references/writing-contract.md 3.6 KB
- requirements.txt 100 B
- scripts/check_deps.py 5.9 KB runs code
- scripts/compile_doc.py 13 KB runs code
- scripts/dot_wrapper.mjs 1.5 KB runs code
- scripts/install_tools.py 7.2 KB runs code
- scripts/render_doc.py 4.2 KB runs code
- scripts/render_dot.mjs 699 B runs code
- scripts/render_visual.mjs 1.6 KB runs code
- scripts/sgw_common.py 1.8 KB runs code
- scripts/validate_evidence.py 11 KB runs code
- scripts/validate_presentation.py 23 KB runs code
- tests/test_literal_rendering.py 4.3 KB runs code
- tests/test_rendering.py 17 KB runs code
- tests/test_validation.py 28 KB runs code
- UPSTREAM.md 2.0 KB
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.
- 4d ago First seen · 128 lines · 30 tokens per session scan A eded01238aff
source-grounded-writing is a skill published in the GitHub repository davidomil/cloudx (1 stars, last pushed 2d ago), licensed MIT. It adds 30 tokens to every session and 2,256 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 2026-11-03.
source-grounded-writing compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| source-grounded-writingdavidomil/cloudx | Skill | 1 | 2d ago | MIT | A | 30 | Claude Code, Codex |
| batch-apiQwenLM/qwen-code | Skill | 29k | yesterday | Apache-2.0 | A | 51 | Claude Code |
| repo-hygieneQwenLM/qwen-code | Skill | 29k | yesterday | Apache-2.0 | A | 48 | Claude Code, Codex |
| test-failure-explorerQwenLM/qwen-code | Skill | 29k | yesterday | Apache-2.0 | A | 32 | Claude Code, Codex |
| coordinateQwenLM/qwen-code | Skill | 29k | yesterday | Apache-2.0 | A | 43 | Claude Code |
| build-teaql-appteaql/teaql-agent-kit | Skill | 2.8k | yesterday | MIT | A | 112 | Codex |
| context-exclusionteaql/teaql-agent-kit | Skill | 2.8k | yesterday | MIT | A | 41 | Claude Code |