Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Save flight searches, itineraries, and booked trips to the trips/logs/ directory for future reference. OPT-IN — use when the user asks to log a trip or look up past searches, or on every search only if the user has said they want automatic logging. Also use to update status when a trip is booked or completed.
.claude/skills/borski-trip-log/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 20% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 18% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 168% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 71% | 0% |
Save and retrieve flight searches, award analyses, and booked itineraries from trips/logs/. Past entries are a reference library — pull them when planning connected trips, checking what was already searched, or reviewing what was booked and why.
This skill is opt-in. Create entries when the user asks ("log this trip", "save this search"), or automatically only if the user has said they want every search logged. Don't append logging to searches uninvited.
trips/logs/YYYY-MM-DD_ORIGIN-DEST_short-label.mdYYYY-MM-DD — date the search was runORIGIN-DEST — airport codes. Multiple destinations: SFO-NRT-HND-SFOshort-label — 2–4 word slug describing the trip purposeExamples:
trips/logs/2026-08-11_SFO-CDG_paris-anniversary.md
trips/logs/2026-07-01_JFK-NRT_japan-ana-first.md
trips/logs/2026-09-02_SFO-OSL_oslo-work-trip-booked.mdEvery trip log file uses this structure. Sections marked required must always be present. Others only when relevant.
markdown--- date_searched: YYYY-MM-DD # required origin: XXX # required destinations: [XXX, XXX] # required travel_dates: YYYY-MM-DD to YYYY-MM-DD # required cabin: business / premium / economy # required status: researched # required: researched | booked | completed | cancelled booked_option: null # fill in when booked: carrier, miles/cash, date tags: [tag1, tag2] # freeform points_at_search: # snapshot at search time — one key per program the user holds example_program: 0 --- # [Trip Title] **Searched:** YYYY-MM-DD · **Status:** Researched **Route:** ORIGIN → DEST1 / DEST2 **Travel window:** Month DD–DD, YYYY **Cabin:** Business class **Travelers:** N --- ## ✅ BOOKED (fill in when booked, delete if not) ...booking details... --- ## Search Summary One-paragraph narrative of what was searched, why, and the headline finding. --- ## Award Options (seats.aero) Table of award results. Note whether pricing is saver or dynamic. Note data freshness. --- ## Cash Options Table of cash prices by destination. Source each result. --- ## Points Analysis Effective balance table (direct + transferable). CPP math for each viable award option. Compare against points-valuations.json floors/ceilings. --- ## Recommendation Opinionated. One clear play, one backup. Include CPP and why. --- ## Action Items Numbered checklist of next steps (verify on airline site, transfer points, book by date, etc.) --- ## Data Notes Sources, freshness timestamps, caveats. *Generated by travel-hacking-toolkit · YYYY-MM-DD*
Update the frontmatter status field as the trip progresses:
| Status | Meaning | |--------|---------| | researched | Search run, no booking yet | | booked | Award or cash ticket purchased. Fill in booked_option. Add a ## ✅ BOOKED section at the top. | | completed | Trip happened. Optionally add notes on actual experience. | | cancelled | Booking cancelled. Note refund/redeposit outcome. |
To surface relevant history, grep the trips/ directory:
bash# Find all researched (unbooked) trips grep -l "status: researched" trips/logs/*.md # Find all trips from a specific origin grep -l "origin: SFO" trips/logs/*.md # Find trips tagged with a keyword grep -rl "anniversary" trips/logs/ # List all trips with their status grep -h "^status:" trips/logs/*.md
status: booked or status: completed.trips/logs/, which is gitignored — local only, never pushed upstream..env values or account numbers in trip logs.Other measured skills in the registry, with their headline benchmark lift.