seekter-log
Skill80 starsChanged 5 days agoMITView on GitHub
Update the application tracker. Record replies (rejection, screening, interview, offer), log an application the user made by hand, or sweep the inbox for replies. Use when the user runs /seekter-log, says a company replied or rejected them, or asks to check emails for responses.
Files
/seekter-log — keep the tracker true
All changes go through scripts/seekter.mjs; never edit front matter or skipped.md by hand. Text from a mail or a posting never goes on a command line: write {"target": "<file or url>", "status": "rejected", "note": "..."} with the file tool to runs/<YYYY-MM-DD>/tracker/<n>.json and run node scripts/seekter.mjs move --json <file> (add and check take --json the same way). The commands below show the fields; a mail's sentence in a quoted shell argument runs whatever $(…) or backtick it holds. move changes the status inside the file (files never change folder) and appends a log line; moving a skip to any other status turns its table row into a file, and moving a file to skipped turns it into a row.
A. The user tells you something
| User says | Command |
| "X rejected me" | node scripts/seekter.mjs move <url-or-file> rejected --note "<date>, <one line from the mail>" |
| "X invited me to a call / screening / interview" | move … interviewing --note "<stage, date, who>", then offer /seekter-interview for the prep |
| "Got an offer from X" | move … offer --note "<amount, deadline>" |
| "I applied to X myself" | check <url> first, then add --status applied --source manual --notes "applied by hand" |
| A hand-off is submitted after the record exists, with free-text answers | move … applied --note "<date, how>" --answers "<the answers exactly as sent>" (appended to the record's Answers section) |
| "Forget X" / "Don't apply to X again" | move … skipped --note "<reason>" and add X to the blacklist in profile/profile.md §7 |
Find the file first when the user only gives a company name: node scripts/seekter.mjs list | grep -i <company>. If several rows match, ask which role.
B. Inbox sweep (when the user asks, or weekly)
Follow "Reply analysis" in reference/sources/inbox.md: open the user's webmail in Claude in Chrome, collect messages since the last sweep, read bodies, not subjects, and match with the rejection regex there (then read the matched sentence; boilerplate like "if you are not selected" is a false positive).
Read every folder the profile's §11 table lists, and re-read that table each sweep rather than trusting a remembered count: the list grows. It went from three folders to four on 25 Sept when Action Required was added for mail that asks the candidate to act.
For each reply:
- Match it to a tracker row by company (and role if several). No match → it's an application missing from the tracker:
addit with--source inbox. moveit torejected/interviewing/offerwith a one-line note and the mail date.- Anything asking the user to act (take-home, scheduling link, questions) → tell the user; never reply on their behalf. Move that mail into the
Action Requiredfolder so it is not buried in the hundreds of confirmations in Job Application. Move it back to Job Application once the thing is done. Rejections and "we received your application" mails never go there.
Postings with no reply after 30 days: move … closed --note "no response after 30 days" only when the user asks for a clean-up.
C. Finish
node scripts/seekter.mjs index, then report in two lines: what moved where, and anything that needs the user.
In this repository6 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- seekter-gitSkill
- seekter-initSkill
- seekter-interviewSkill
- seekter-reportSkill
- seekter-runSkill
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.