travel-agent
CLAUDE.md1 starsChanged 3 months agoMITView on GitHub
This repo is a travel-planning method (published frameworks + playbooks) and an agent that runs it. Read _start_here.md first — it's the 60-second orientation and the map of what's in here. The plan-trip skill is the main entry point. Almost anything that looks like "help me plan a trip" should invoke it rather than being answered freehand — its PROTOCOL.md (.claude/skills/plan-trip/PROTOCOL.md) encodes the intake → architecture → itinerary → calibration → gear pipeline, each stage pulling its method from a…
In this repositoryOne other file it gives its agents.
- plan-tripSkill
The file
We split this CLAUDE.md into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 27
trips/is where the work happens: Every trip the skill plans lives at trips/<destination-slug>-<duration>d/trip.md, scaffolded from … - Distinctive: line 49
Never add a "Co-Authored-By: Claude" (or any Claude/Anthropic) trailer to commit messages, PR descriptions, or any git-related output for …
- Very specific: line 55
If you fork this repo and commit your own trips, trips/ is not scrubbed for you — keep real personal data (passport numbers, exact …
≈ “Common” is an estimate: at least 1 in 5 of the 28,029 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Claude working rules — travel-agent
This repo is a travel-planning method (published frameworks + playbooks) and
an agent that runs it. Read _start_here.md first — it's the 60-second
orientation and the map of what's in here.
Entry point
The plan-trip skill is the main entry point. Almost anything that looks
like "help me plan a trip" should invoke it rather than being answered
freehand — its PROTOCOL.md (.claude/skills/plan-trip/PROTOCOL.md) encodes
the intake → architecture → itinerary → calibration → gear pipeline, each
stage pulling its method from a specific framework or playbook file instead of
generic travel advice. Read SKILL.md first for the pointer, then
PROTOCOL.md for the actual flow before running it.
Read-only vs. editable
frameworks/andplaybooks/are read-only reference. They're the published methodology this repo teaches — hub-and-spoke architecture, the 8-day novelty peak, jet-lag protocols, threat tiers, capsule wardrobes, gear kits, points strategy. Users fork this repo to use the method, not to rewrite it. Don't edit files in these directories while runningplan-tripor answering a planning question.trips/is where the work happens. Every trip the skill plans lives attrips/<destination-slug>-<duration>d/trip.md, scaffolded fromtrips/_template/trip.md. This is the only directoryplan-tripwrites to.
The pipeline
- Intake — a short questionnaire, batched, asking only what wasn't already given.
- Architecture — hub-and-spoke vs. circuit, decided from
frameworks/trip_architecture.md. - Itinerary — a day skeleton respecting the 8-day novelty peak, with
frameworks/jet_lag_protocol.mdapplied to arrival days. - Calibration — destination threat tier from
frameworks/situational_awareness.md, posture and digital-axis notes. - Gear — a delta checklist against
playbooks/clothing_capsule.mdandplaybooks/gear_kit.md, not a full manifest.
Full detail lives in .claude/skills/plan-trip/PROTOCOL.md — this section is
the map, not the spec.
Git commit attribution
Never add a "Co-Authored-By: Claude" (or any Claude/Anthropic) trailer to commit messages, PR descriptions, or any git-related output for this repo. Commit as the repo owner only.
Forking this publicly
If you fork this repo and commit your own trips, trips/ is not scrubbed for
you — keep real personal data (passport numbers, exact addresses, booking
confirmations, anything you wouldn't want public) out of anything you commit.
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.

