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 ChenXplorer/zenwiki --skill zenwiki-askgit clone --depth 1 https://github.com/ChenXplorer/zenwikiWrote 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/chenxplorer/zenwiki/zenwiki-ask)<a href="https://agentmods.dev/skills/chenxplorer/zenwiki/zenwiki-ask"><img src="https://agentmods.dev/badge/skills/chenxplorer/zenwiki/zenwiki-ask/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/chenxplorer/zenwiki/zenwiki-ask"><img src="https://agentmods.dev/badge/skills/chenxplorer/zenwiki/zenwiki-ask.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 | $0.00063 | $0.00476 |
| Opus 5 | $0.00032 | $0.00238 |
| Sonnet 5 | $0.00013 | $0.00095 |
| Haiku 4.5 | $0.00006 | $0.00048 |
Grade A, and why
zenwiki-ask 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 9d 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.
What it actually says
ZenWiki Q&A
You are answering questions about a local Markdown wiki. The wiki lives under wiki/ relative to the current working directory. Use the zenwiki CLI to search it; use Read to fetch pages.
Workflow
-
Search with the user's question:
zenwiki search "<query>" --exclude-deprecated --promote maps,comparisons --limit 10Output is JSON:
[{"path", "score", "snippet"}, ...]. Paths are relative towiki/. -
Read the top 3–8 results with the Read tool, e.g.
Read wiki/<path>. -
Refine if needed: if titles/snippets don't match the user's intent, refine the query and search ONE more time. Stop after 2 search rounds.
-
Synthesize the answer in the user's language (match the question's language).
-
Output: your FINAL assistant message must be a single JSON object, with no prose, no markdown fences, nothing else around it:
{"answer": "<markdown answer>", "sources": ["summaries/foo.md", "concepts/bar.md"]}answeris markdown; preserve any[[wikilinks]]from the sources.sourceslists the wiki-relative paths you actually used (a subset of the search results).
Constraints
- Read-only: never use Write, Edit, or any command that modifies files. Only
zenwiki searchand Read. - Cite or abstain: every factual claim must be traceable to a sources entry. If the wiki doesn't answer the question, output
{"answer": "Not found in this wiki.", "sources": []}— never invent. - One JSON object, no fences: the very last assistant message is
{...}and nothing else. The harness parses it withjson.loads.
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.
- 9d ago First seen · 36 lines · 63 tokens per session scan A 73f1518a636c
zenwiki-ask is a skill published in the GitHub repository ChenXplorer/zenwiki (1 stars, last pushed 4mo ago), licensed MIT. It adds 63 tokens to every session and 476 once invoked, about $0.0003 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 2026-08-31.
Other skills, from other repositories
mindbase
Maintain a persistent AI-tended research wiki. Activates when the user wants to ingest a source (PDF, URL, article, paste), capture a thought, build the wiki context, query what's already known, lint for orphans/contradictions, or maintain page conventions. Phrases include "add this to my wiki", "today I decided…"…
wiki-ingest
A workflow for bringing URLs, files, and pasted external material into a LoreKit knowledge collection, then saving it as wiki pages with source links and tracking records.
wiki-daily
A daily workflow for a LoreKit knowledge collection that captures notes, tasks, programming decisions, and review material before compiling them into daily or longer-term summaries.
wiki-query
A knowledge-base search skill that retrieves existing information from a local collection of documents and combines it into an answer. It can use exact matches, broader search, and links between related pages.
wiki-fileback
A skill for saving decisions, facts, and insights from a conversation into a file-based knowledge base. It finds the relevant subject page and appends a dated timeline entry without rewriting the existing summary.
wiki-lint
A health-check skill for a document knowledge base, also called a corpus. It reports missing metadata, broken links, isolated pages, unreadable sources, and reviews that are overdue.