job-recap

Job Recap

Weekly retrospective for the job search. Reads every data source the pipeline writes to, slices it by the requested time range, and produces a single Friday-evening recap: numbers, wins, risks, and 3 concrete moves for next week.

Where /job-dashboard answers "what does the pipeline look like right now" and /job-patterns answers "why is the pipeline shaped that way", /job-recap answers "what actually happened in the last 7 days, and what should I do next week."

Repo location

$AI_JOB_AGENT_ROOT → ~/.claude/skills/ai-job-agent/ → REPO_PATH marker file → ~/ai-job-agent/.

Status emoji (consistent with /job-track and /job-patterns)

📄 applied · 📬 submitted · 💼 interview · 🎯 offer · ❌ rejected · 🚫 blocked · 🚪 withdrawn

Workflow

Step 1 — Resolve repo + parse the time range

1. Resolve $AI_JOB_AGENT_ROOT (env var → default skills path → REPO_PATH marker → ~/ai-job-agent/).
2. Parse $ARGUMENTS:
   • empty or 7d → start = today - 7 days
   • 14d → start = today - 14 days
   • 30d → start = today - 30 days
   • since YYYY-MM-DD → start = that date
   • anything else → echo the arg, default to 7d, tell the user
3. Compute prior_start = start - (today - start) so we can show a delta against the prior window of equal length.
4. Print one line: "Recap window: <start> → <today> (N days). Prior window for delta: <prior_start> → <start>."

Step 2 — Read all the data sources

Read in parallel; treat each missing/empty source as 0 (don't fabricate):

1. application-tracker.csv — filter to date ≥ start. Columns: date, company, role, status, location, source, applied_by, url, notes, contact, compensation, days_since, key.
2. outreach-log.csv — filter to sent_at ≥ start and separately collect rows where last_follow_up_at ≥ start (so a follow-up sent this week to an old contact still counts as activity).
3. reports/*.md — Glob the directory. Parse YAML frontmatter and filter by evaluated_at ≥ start. Pull out fit_score if present.
4. interview-prep/*.md — Glob, filter by file mtime ≥ start.
5. output/*.pdf — Glob, filter by mtime ≥ start. These are CV PDFs from /job-cv and /job-evaluate.
6. Git log — try git -C $AI_JOB_AGENT_ROOT log --since="$start" --oneline. If the repo is not a git repo (fatal: not a git repository), treat the result as empty and don't error.
7. config/search-plan.md — if it exists, read the file and pull the "Log" section's entries within range. Also note the file's mtime — if today - mtime > 14d, flag the plan as stale.