li-audit

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: the skill runs on data I supply myself about my own LinkedIn posts — either the analytics export CSV I download myself from my own LinkedIn account (LinkedIn: Analytics -> Content -> Export), or a screenshot per post (impressions, reactions, comments, reposts), or simply my past posts with their reaction counts. Getting that export, if I choose that route, means I log into LinkedIn myself: this skill never logs into LinkedIn, never asks for my LinkedIn password, cookies or session, uses no browser automation and no LinkedIn API, and posts nothing. There is no script and no code in this skill at all: it is a single instruction file. Install "li-audit" for me. It is a post-mortem on what I have already published on LinkedIn: from the analytics or past posts I supply, it computes engagement rate ((reactions + comments + reposts) / impressions), comment ratio (comments / reactions), reach multiple (impressions / follower count) and, if available, save/send rate, ranks my posts by engagement rate and reach multiple rather than by raw impressions, puts the top 5 and bottom 5 side by side, looks for what separates them (hook formula, format, length, theme, day and time checked last, first-hour comments), and states each finding as a claim with its evidence and a confidence note (with about 30 posts a pattern can be seen, with 6 it cannot), ending in a STOP / DO MORE audit block in the format shown in SKILL.md. Repository: https://github.com/avi691/linkedin-agent-skill (the skill is the whole skills/li-audit/ folder at the pinned commit: SKILL.md only, 1 file) 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-audit/ 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 bdb5b12c699ca5d512b785f7168f8e494e020a1ff1195aaa9a58cd5b193c1bf2 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, 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 file: - The skill folder holds 1 file only: SKILL.md (3,417 bytes). There are no scripts, no code, no JSON data files and nothing executable in the folder, and nothing to run at install time or at use time. - The references outside the folder, exactly as written in SKILL.md, and what they do and do not require: (a) at use time it also reads ~/.claude/linkedin/log.md if that file exists — a local file, written by other skills in the same pack (li-post appends each post there with its date and hook), which records which hook formula each of my posts used; reading it is optional (SKILL.md says "if it exists"), it is not part of this install, and you must not create it or install another skill to supply it. (b) The pattern step asks which numbers from hooks.json are in the top 5 — hooks.json is not in this folder: in the repository it is skills/li-post/hooks.json, the 21 hook formulas of the separate li-post skill. Without li-post, that one check is limited to the hook numbers or names already recorded in log.md or supplied by me; every other metric and pattern check works from my supplied data alone. Do not install li-post or any other skill as part of this install. (c) At the end, SKILL.md says to hand the conclusions to /li-plan (the pack's weekly-planning skill) so next week's plan is built on them — that is an optional hand-off to another skill, not a file, not a service, and not a step of this install; if li-plan is not installed, simply report the conclusions to me. Apart from these three points, SKILL.md needs no file from elsewhere in the repository — not the other 10 skills, not templates/voice.md (SKILL.md never mentions it, even though the repository README says every skill in the pack reads a voice profile), and not the README. - 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, password, cookie or account is asked for, and there is no telemetry in the folder — it contains no URL at all. - What happens at use time, disclosed here before it ever happens — using the skill is a separate step that only starts when I ask for an audit and supply my data: the skill itself makes no network call, calls no API, opens no browser, does not log into LinkedIn, does not read LinkedIn cookies or any session store, and posts, schedules and sends nothing anywhere. The only things that enter our conversation are what I choose to give you (my LinkedIn analytics CSV export, screenshots of my posts, or my past posts and their counts, which I obtained from my own LinkedIn account myself) and, if it exists, the contents of my local ~/.claude/linkedin/log.md, which you read locally so the hook-formula check can use it. As with anything I paste or you read into this conversation, that material is processed by you, my agent, under your own account and terms; the skill sends none of it to LinkedIn or to any third-party service, and there is no other recipient named in SKILL.md. - Do not ask for my LinkedIn password, cookies or session, do not log into or scrape LinkedIn, do not install li-post, li-plan or any other skill, and do not run any audit during this install: an audit only starts later, when I supply my analytics or past posts and ask for it. 3. When I say "yes, install", install it as a skill: copy the whole li-audit folder into the agent's skills directory, in a folder named "li-audit", keeping its 1 file (SKILL.md at its root). A manual clone followed by copying the skills/li-audit 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-audit), the 1 file in place with matching fingerprint. There is no script to run to verify. 5. Report what was installed, where, and what I still need to do myself (for example: supply the audit input myself when I want an audit — download the analytics CSV export from my own LinkedIn account myself, or provide a screenshot per post, or paste my posts with their reaction counts, and tell you my follower count if I want reach multiples; optionally keep ~/.claude/linkedin/log.md if I use the other pack skills that write it, knowing li-audit only reads it if it exists; 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.