Obsidian Wiki

v1.0.2

Build and maintain a compiled LLM Wiki inside an Obsidian vault following Karpathy's pattern. Use when working with an Obsidian vault as a persistent knowled...

0· 4·0 current·0 all-time
byJérôme Benoit@jerome-benoit
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name/description match the provided artifacts. The scripts perform vault-centric tasks (ingest, manifest/hash tracking, index generation, linting, wikilink fixes) that are expected for an 'Obsidian Wiki' compiler; no unrelated binaries or credentials are requested.
Instruction Scope
SKILL.md instructs the agent to read Obsidian config to locate the vault and to read/write files under the vault. This is appropriate for the stated purpose but is file-system access that can modify user data (wiki/ and .wiki-meta/ are explicitly written). The README asks for user confirmation when the vault path is inferred; follow that guidance and prefer dry-run modes when available.
Install Mechanism
Instruction-only skill with bundled scripts; there is no download/install step. Scripts rely on common system tools (bash, python3, pdftotext optional). No external code fetches or archives are present.
Credentials
No environment variables, credentials, or external endpoints are required. The scripts operate on files under the provided vault path only (they do read Obsidian config files to discover a vault path, which is consistent with their purpose).
Persistence & Privilege
always:false (not force-included). disable-model-invocation:false (agent may invoke autonomously) — this is the platform default. Because the skill writes to the user's vault, consider whether you want it to run autonomously; otherwise require user confirmation before write operations as recommended by SKILL.md.
Assessment
What to consider before installing: - This skill will read your Obsidian config to locate a vault and will read and modify files under the vault (it creates/updates files in wiki/, writes a manifest to .wiki-meta/manifest.json, and may update _meta/ after approval). That behavior is expected but means the skill can change your notes — back up the vault before first run. - Scripts are local Bash/Python programs (no network calls or credential exfiltration detected). Review the included scripts (fix-wikilinks.py, wiki-manifest.sh, wiki-lint.sh, wiki-index.sh, extract-book-digests.sh) yourself and run them in --dry-run modes where supported before allowing write actions. - The skill suggests confirming with the user if the vault path is inferred; enforce that confirmation. Pay attention to the fix-wikilinks tool which will rewrite many files — run dry-run first. - If you are uncomfortable with autonomous agent invocation (the agent can call the skill automatically), restrict its invocation or require manual triggers. - If you keep sensitive documents in the vault/raw directory, consider moving them out or ensuring you trust the agent, since the skill reads raw/ sources for ingestion (there is no network transfer in the scripts, but the agent could include content in model prompts during processing).

Like a lobster shell, security has layers — review code before you run it.

latestvk97382gt67yz3m5yvbg5ds0bgd84j26c

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Comments