trail
Skill1 starsChanged 6 months agoMITView on GitHub
Trail is a local web UI that visualizes what Claude Code has been doing — sessions, tool calls, file activity, memory, subagents — for every Claude Code project on the user's machine. Invoke when the user says any of "open trail", "start trail", "show me trail", "trail aç" (Turkish), "show me what claude did", "show claude's path", "see my claude sessions", "show me the memory", "open the project in browser", "trail studio aç", or "vault studio aç" (legacy name), or otherwise asks to inspect/visualize their Claude Code activity. Runs at http://localhost:7777, 127.0.0.1-only, no LLM calls.
- Pipes a download into a shell
- Reads credentials
5 other projects use a version of this.
Adopted by
Most recently taken up by
- Namann-14/trailtheir own version, first seen Nov 15, 2026
- ntholm86/principles-of-earned-autonomy-skills-suitetheir own version, first seen Nov 11, 2026
- 5throck/ai-workspace-standardstheir own version, first seen Nov 1, 2026
- simota/agent-skillstheir own version, first seen Oct 23, 2026
- MichaelTarasov02/Trailtheir own version, first seen Oct 20, 2026
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- simota/agent-skills★ 99same name
- 5throck/ai-workspace-standards★ 25same name
- ntholm86/principles-of-earned-autonomy-skills-suite★ 4same name
- MichaelTarasov02/Trail★ 0same name
- Namann-14/trail★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
We split this skill into the rules it gives an agent
5 rules. 0 are common, 2 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 5 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 8
When the user wants to open Trail, run the Bash script below.
- Distinctive: line 10
After it boots, tell the user the URL and how to use it (see "After starting" below).
- Distinctive: line 45
Default language is English.
- Very specific: line 46
Stop with Ctrl+C in the terminal.
- Very specific: line 76
Bun >= 1.0: must be installed. If missing, the launcher prints curl -fsSL https://bun.sh/install | bash.
“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Trail Skill
When the user wants to open Trail, run the Bash script below. It resolves the plugin root in 4 fallback steps so it works regardless of how the plugin was installed (marketplace, dev symlink, or via $CLAUDE_PLUGIN_ROOT).
After it boots, tell the user the URL and how to use it (see "After starting" below).
PLUGIN_DIR="${CLAUDE_PLUGIN_ROOT:-}"
if [ -z "$PLUGIN_DIR" ]; then
CANDIDATE=$(ls -dt "$HOME"/.claude/plugins/cache/*/trail/*/scripts 2>/dev/null | head -1)
[ -n "$CANDIDATE" ] && PLUGIN_DIR=$(dirname "$CANDIDATE")
fi
if [ -z "$PLUGIN_DIR" ]; then
CANDIDATE=$(ls -dt "$HOME"/.claude/plugins/marketplaces/*/scripts 2>/dev/null | head -1)
[ -n "$CANDIDATE" ] && PLUGIN_DIR=$(dirname "$CANDIDATE")
fi
if [ -z "$PLUGIN_DIR" ] && [ -L "$HOME/.claude/skills/trail" ]; then
RESOLVED=$(readlink "$HOME/.claude/skills/trail")
PLUGIN_DIR=$(cd "$(dirname "$RESOLVED")/.." && pwd 2>/dev/null)
fi
if [ -z "$PLUGIN_DIR" ] || [ ! -d "$PLUGIN_DIR/scripts" ]; then
echo "Trail: plugin path not found. Try /plugin marketplace list."
exit 1
fi
bash "$PLUGIN_DIR/scripts/start.sh"After starting
Tell the user:
- The URL the script printed (typically
http://localhost:7777). - They can switch projects from the top-left dropdown — Trail auto-discovered every Claude Code project on this machine.
- The first visit shows a Welcome overlay explaining the 7 tabs (Notes, Files, Trail, Memory, Agents, History, Guide). They can dismiss it or start with the Guide.
- Default language is English; an EN/TR toggle is top-right if they prefer Turkish.
- Stop with
Ctrl+Cin the terminal.
Trigger examples (English)
- "open trail"
- "start trail"
- "show me what claude did"
- "show me my claude sessions"
- "open the project in browser"
- "I want to see claude's path"
- "show me the memory"
Trigger examples (Turkish)
- "trail aç"
- "claude'un izini göster"
- "trail studio aç"
- "vault studio aç" (legacy / earlier name)
What Trail is (so Claude can explain it accurately)
- Pure local visualization — no LLM calls, no data leaves the machine.
- Server binds to
127.0.0.1only (LAN/internet not reachable). - Reads from
~/.claude/projects/<encoded-path>/*.jsonl(the user's own Claude Code session logs). - Writes only to the active project's vault — and only
.md/.mdx/.markdown/.txtfiles. - Secret-guarded:
.env*,*credentials*,*api_key*,*.pem,*.key, etc. are masked + readonly. - CSP-locked: no external scripts/styles loaded by the UI.
Requirements
- Bun >= 1.0 must be installed. If missing, the launcher prints
curl -fsSL https://bun.sh/install | bash. - Works on macOS and Linux out of the box. Windows requires WSL or git-bash (bash + standard Unix tools).
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.

