backfill
Skill6 starsChanged 3 days agoMITView on GitHub
Backfill the workspace from the internet with web research — read the product website to fill gaps in the product doc, search HN, Reddit and the web for existing mentions, file already-published posts with their engagement, and stub competitor profiles from public pages. Use when the user says "backfill the workspace", "import what already exists", or right after setup for a product with any public history.
5 other projects use a version of this.
Adopted by
Most recently taken up by
- lucamartins/ouroborostheir own version, first seen Nov 16, 2026
- clins1994/jobcan-agenttheir own version, first seen Nov 8, 2026
- ronniepinnell/caspertheir own version, first seen Nov 8, 2026
- scaccogatto/okf-skillstheir own version, first seen Oct 13, 2026
- chrisselig/claude-code-toolkittheir own version, first seen Oct 9, 2026
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- scaccogatto/okf-skills★ 444same name
- lucamartins/ouroboros★ 2same name
- chrisselig/claude-code-toolkit★ 1same name
- clins1994/jobcan-agent★ 1same name
- ronniepinnell/casper★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
We split this skill into the rules it gives an agent
2 rules. 0 are common, 2 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 4 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 27
Mention sweep: Search the product name + aliases: HN (Algolia search), Reddit, general web and blogs.
- Distinctive: line 46
Replies already posted: If the user has answered people in others' threads, ask for the links.
“Common” is measured: kept by at least 6 in 10 of the 4 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Backfill from the internet
The user's marketing didn't start today. This skill imports the campaign history that already exists — mentions, posts, competitor facts, baselines — so the graph starts true, not empty. Requires web access. Collaboration over automation: propose before filing in bulk, and never state what a page didn't show.
Steps
- Scope with the user. Confirm: the product name and searchable aliases
(old names, handles, the domain), the website URL, competitor names, and
which platforms they've already posted on (handles or profile links).
Ideally run after setup; if
data/product.mdis still unfilled, the website pass may propose drafts for its gap sections, clearly marked for the user to confirm. - Website pass. Fetch the site. Fill gaps only:
data/product.mdsections still showing ✏️, pricing intodata/offer.md, the page inventory intodata/website.md(Key pages), org facts where visible. Where the site contradicts something the founder wrote, ask — never silently overwrite. - Mention sweep. Search the product name + aliases: HN (Algolia search),
Reddit, general web and blogs. Open every hit before filing (Reddit blocks
most server-side fetches — a thread that refuses to load usually works via
https://embed.reddit.com/<post-path>). Atomic comment/thread →data/mentions/YYYY-MM-DD-<slug>.md(verbatimquote, directresource,authorexactly as displayed, honestsentiment); article-length piece →data/external-<slug>.mdwith a one-linetakeaway. Link mentions fromdata/mentions.mdand the venue's## Mentionsheading; file authors worth following up indata/people/(roles via the role hubs) — a positive mention is warmed-up outreach. - Published-post import. From the user's links and their profile pages,
file each already-published post as
data/posts/<platform>/YYYY-MM-DD-<slug>.md(Reddit nests per subreddit —AGENTS.md, What goes where;stage: published, the realpublisheddate andresource; body = the actual post text, or a faithful summary when it's long). Record engagement numbers as displayed. The schema requiresgoalandsuccess— set them retroactively with the user ("what was this post for?"). Link each post from its venue's## Postsheading and its pillar hub, and flip the venue'sstagetoactive. - Replies already posted. If the user has answered people in others'
threads, ask for the links. File each as the draft skill's reply mode does
(thread page, lead if relevant, reply with the comment verbatim and
originas the user says), withoutcomeread from the thread as it stands andreviewed: <today>. - Competitor snapshots. For each competitor named, fetch their public
pages and fill
data/competitors/<slug>.md: Positioning, Pricing, Strengths and Weaknesses honestly, a dated Changelog entry,updated: <today>. - Baselines. Public numbers found along the way (stars, followers,
subscribers, review counts) →
data/metrics.md, each with a dated snapshot line. - Report, validate, commit. Summarize what was filed (n mentions, n
posts, n replies, n people, n competitor profiles, metrics captured) and what could
not be searched (auth-walled platforms, rate limits, dead links) so the
user can fill those by hand.
iwe schema validatemust pass. In a git repository, commitbackfill: <n> mentions, <n> posts, <n> competitors.
Rules
- Never invent. Quotes verbatim from the fetched page; numbers only as displayed; dates from the source. A find without a URL you actually opened does not get filed.
- Fill gaps, don't overwrite — the founder's language wins; conflicts become questions, not silent edits.
- Propose the find-list before bulk-filing (what, where found, how it would be filed); file after the user confirms. A single obvious find can be filed directly.
- A mention the user remembers but that's gone at the source: file it only with a working archive link, else skip it and say so.
- Degrade honestly: report the platforms you couldn't search rather than implying full coverage.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

