li-profile

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: none for the install itself: no account, no API key, no paid service, and nothing else to install first. For using the skill afterwards, read this first: there is no LinkedIn login, no browser automation, no API and no key — the skill works only on LinkedIn profile text I paste myself (headline, about section, current role and my last two experience entries, plus whether I have a banner and a featured section; a screenshot of the top card is enough for a first pass), and nothing is posted or saved to LinkedIn by the skill — I paste each rewrite in myself. One part of the skill as written depends on a sibling skill that this install does NOT include: its Output step says every rewrite is "already run through /li-human" — /li-human is a separate skill in the same repository (skills/li-human/, with its own Python scripts), not part of this folder; without it, the score and the rewrites still work, but that humanizer pass cannot run as described (see point 2). Also note the pack-level voice file: the repository README and templates/voice.md say every skill in the pack reads ~/.claude/linkedin/voice.md, but li-profile's own SKILL.md never mentions voice.md — this install does not copy that template and does not create that file. Install "li-profile" for me. It scores a LinkedIn profile out of 100 against the 12-part rubric in its bundled rubric.json (headline 12, about first 2 lines 10, about body 10, featured 8, banner 6, photo 6, current role 10, experience depth 8, skills 6, recommendations 8, activity 10, contact 6), shows the per-item table and the total, then rewrites the parts that lose points in fix-first order (descending points lost, not everything at once): headline (220 characters, three options, formula {what you do for whom} | {proof} | {how to start}), about first two lines, about body (second person, under 1,400 characters), featured (three items: best post, proof asset, way to contact), experience (one scope line plus two to three outcome bullets with numbers per role, anything older than ten years cut to a single line), and banner (one sentence of positioning and one way to reach me), then re-scores and shows the delta honestly. Repository: https://github.com/avi691/linkedin-agent-skill (the skill is the whole skills/li-profile/ folder at the pinned commit: SKILL.md plus rubric.json, 2 files) Pinned version: install the code at commit e4ec4fb6f1b0e227eb9b14b31b735072db5df9cb exactly. Do not substitute a newer version. If this commit is unreachable, stop and tell me. 1. Clone https://github.com/avi691/linkedin-agent-skill into a temporary folder, check out commit e4ec4fb6f1b0e227eb9b14b31b735072db5df9cb, and copy the skills/li-profile/ folder from that checkout. Then compare every file in the folder with the fingerprints recorded for this listing on 2026-10-19: [FINGERPRINTS] - SKILL.md: SHA-256 4f0ebdcf8a3ef035d70aae89838ff916fc8a33a4a12b333e58bb2bc41727d11e - rubric.json: SHA-256 d1add490dad3cc32aab4826c3d5a40bd1552d26a92424deaad9ee2e3c7c05d14 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 read rubric.json in full, and tell me in plain words: what the skill does when used, which services it connects to, whether it reads or sends any personal information or keys, 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 2 files only: SKILL.md (3,271 bytes) and rubric.json (2,462 bytes). There is no code, no script and no executable in this folder: the skill is instructions plus a scoring rubric. rubric.json is version 1.0, twelve items totalling 100 points, each with what full marks looks like. - The skill is self-contained for scoring and rewriting, with exactly two documented links outside its folder, and neither is installed or created by this install: (a) the Output section of SKILL.md says each rewrite is "already run through /li-human" — /li-human is a sibling skill at skills/li-human/ in the same repository (its own SKILL.md plus humanize.py, detect.py and slop.json); it is not in this folder, so if I have not installed it separately, that humanizer pass cannot run and the rewrites are plain rewrites — do not install li-human or any other li-* skill as part of this install. (b) The repository README and templates/voice.md say every skill in the pack reads ~/.claude/linkedin/voice.md (a voice profile the user copies from templates/voice.md and fills in), but li-profile's own SKILL.md never mentions voice.md or that path — do not copy templates/voice.md and do not create ~/.claude/linkedin/voice.md during this install. - Nothing is sent anywhere during installation itself; the only download at install time is the skill folder from GitHub at the pinned commit. No script is run during installation, no key is asked for, and there is no telemetry in the folder. The skill contains no access to LinkedIn cookies or session, no credential reading, and no browser automation: SKILL.md says explicitly "Do not log into LinkedIn on their behalf" and "Nothing is saved to LinkedIn by this skill. The user pastes each section in." - What leaves this machine at use time, disclosed here before it ever happens — using the skill is a separate step that only starts when I ask to score or fix a profile: the skill itself makes no network call, calls no API, opens no browser and reads no profile on LinkedIn. I paste my own profile content into our conversation myself (headline, about section, current role and last two experience entries, whether I have a banner and a featured section, and optionally a screenshot of the top card for a first pass), and that pasted text or screenshot is handled by you, my agent, under my own account as part of our conversation — the skill does not send it anywhere else. The scoring (reading rubric.json and scoring each of the twelve items) and the rewriting stay in our conversation and produce copy-ready blocks in the chat. The only way a rewrite reaches LinkedIn is that I paste it into LinkedIn myself, by my own action, outside the skill. The skill never reads a third party's profile directly: it works only on the profile text I supply. - Do not log into LinkedIn, do not open LinkedIn on my behalf, do not install /li-human or any other sibling skill, do not copy templates/voice.md or create a voice file, and do not score or rewrite any profile during this install: setup of those separate pieces is my decision, later, when I actually ask to use the skill. 3. When I say "yes, install", install it as a skill: copy the whole li-profile folder into the agent's skills directory, in a folder named "li-profile", keeping all 2 files (SKILL.md and rubric.json at its root). A manual clone followed by copying the skills/li-profile folder into ~/.claude/skills/ is the route described here; any other route is fine only as long as every installed file matches its fingerprint. 4. Verify with no network calls: frontmatter valid (name: li-profile), all 2 files in place with matching fingerprints. Do not score a profile to verify. 5. Report what was installed, where, and what I still need to do myself (for example: paste my own headline, about section, current role and last two experience entries, plus my banner/featured status — or a screenshot of the top card — when I want a score; decide separately whether to install the sibling /li-human skill if I want the rewrites run through its humanizer pass exactly as SKILL.md describes; decide separately whether to fill in the pack's voice template at ~/.claude/linkedin/voice.md, knowing li-profile's own SKILL.md does not mention it; paste each rewrite into LinkedIn myself, since the skill saves nothing there; using the skill is a separate step that only starts when I ask for it, and the use-time handling disclosed above happens only then). 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.