Defuddle

Before you start: read this whole install prompt, tell me in a few lines what this skill does and what installing it involves, and wait for my go-ahead before you do anything else. Prerequisites — USE REQUIREMENTS DISCLOSED FIRST: this skill works on your own Obsidian vault and runs through the portable core of the claude-obsidian repository, scripts/claude-obsidian.py, which lives at the repository root OUTSIDE this skill's folder, together with the shared contracts it links to (under skills/wiki/references/ in that same product root, for example the provenance and operation-transaction contracts). Installing this listing copies only this skill's folder: without a full claude-obsidian installation (the core, its contracts, python3 to run it, and a selected user vault) the skill cannot run its transactions as designed. The family orchestrator is the separate wiki skill of the same repository. No account, no key and no payment are needed to copy the files, and nothing is run during installation. HONEST LIMIT, STATED IN THE SKILL ITSELF: despite its name, this package performs NO web extraction and NO network fetch. Network extraction remains unavailable until an enforcing adapter is verified; what it produces for a URL is an inert plan (execute: false). The upstream Defuddle CLI is an optional external tool that must NOT be run under this workflow. Install "Defuddle" for me. It plans optional Defuddle extraction of a web page into clean Markdown, and retains already available local source files in your vault, but execution stops at the plan: for an HTTPS URL it builds an inert plan through the portable core (capture external-plan) that executes no network request, reports the normalized host, the required network egress, the redirect policy and the optional dependency, and is marked execute: false. URLs with credentials, fragments, sensitive query parameters or non-public hosts are rejected, and a URL is never interpolated into a shell string. The reason, stated in the skill: the upstream Defuddle CLI it inspected follows redirects and enables asynchronous extraction, so even a local HTML file can trigger extra requests or third-party fallbacks that consent for the initial URL cannot cover; the bare CLI must therefore not be run against a URL or local HTML under this workflow, and no browser, fetcher, proxy or cookie substitute may be silently used. What it does do today: when you ask to retain an existing local HTML or Markdown file, you place it in your vault's inbox/ first, and the reviewed capture workflow (capture plan, then capture apply with the approval hash, APPROVAL_SHA256 pattern) preserves its bytes, create-only, into .raw/captured/ storage; that capture does not clean HTML, verify metadata or synthesize pages. It installs no tools and claims no token savings; cleaning, raw capture and wiki ingestion are three separate operations. Repository: https://github.com/AgriciDaniel/claude-obsidian (the skill is the whole skills/defuddle/ folder at the pinned commit: SKILL.md only, 1 file) Pinned version: install the code at commit 2e545563172f97f02ecaaaab376ceea1246a0583 exactly. Do not substitute a newer version. If this commit is unreachable, stop and tell me. 1. Clone https://github.com/AgriciDaniel/claude-obsidian into a temporary folder, check out commit 2e545563172f97f02ecaaaab376ceea1246a0583, and copy the skills/defuddle/ folder from that checkout. Then compare every file in the folder with the fingerprints recorded for this listing on 2026-10-22: [FINGERPRINTS] - SKILL.md: SHA-256 35958e30b3521ee04155e06253713871c11d82ee840271d9bb091f5c54efaeca If a file does not match, it changed since the listing was prepared: do not install it yet, tell me what you found and let me decide. 2. Read SKILL.md in full, and tell me in plain words: what the skill does when used, what it connects to, whether it reads or sends any personal information or credentials, and anything that seems unusual. Stop on any red flag and tell me before going further. Confirm these points from the files: - The skill folder holds 1 file only: SKILL.md. There is no script, no extractor and no bundled Defuddle CLI in it: the core it calls (scripts/claude-obsidian.py) lives outside this folder, in the full claude-obsidian product root, and the Defuddle CLI is an optional external tool that this workflow forbids running. - Nothing is sent anywhere during installation itself; the only download at install time is the skill folder from GitHub at the pinned commit, and no URL is planned, fetched or cleaned. - At use time the URL path is an inert plan only: it executes no network request (execute: false), so nothing is fetched from any website by this skill. A discovered Defuddle executable is reported as configured, never run merely because it was found and never relabelled verified. - The one real vault write at use time is the reviewed filesystem capture: an existing local file you placed in the vault's inbox/ is preserved byte-for-byte, create-only, into .raw/captured/, via capture apply gated by the approval hash (APPROVAL_SHA256 pattern). It does not clean the HTML. - Do not plan or fetch any URL, do not run any Defuddle CLI or substitute fetcher, and do not capture any file during this install: using it is a separate step that only starts when I ask for it later. 3. When I say "yes, install", install it as a skill: copy the whole defuddle folder into the agent's skills directory, in a folder named exactly "defuddle", keeping its single file (SKILL.md) at its root. 4. Verify with no network calls: frontmatter valid (name: defuddle), the 1 file in place with matching fingerprints. Do not run anything from the folder and do not use the skill to verify. 5. Report what was installed, where, and what I still need to do myself (for example: that it performs no web extraction or fetch at all (URL handling stops at an inert, reviewed plan), that a full claude-obsidian installation and my own Obsidian vault are required to use its capture workflow, and that the optional external Defuddle CLI must not be run under this workflow). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.