Context Timeline

Hooks monitoring 1,164 Scanned by NVIDIA SkillSpector

Real-time browser visualization of Claude Code's context window and subagent/tool execution as a git-graph timeline. Opens http://localhost:7878 on SessionStart with a vertical timeline (most recent on top), one column per agent (main + subagents branching from their Task tool call). Tool calls are color-coded by type (Read=green, Edit/Write=orange, Bash=red, Grep/Glob=cyan, Task=purple, Web=yellow, MCP=gray). Right sidebar shows context-window usage (tokens/200K) with cache_read/cache_creation/input/output breakdown plus per-subagent mini-context. Reads ~/.claude/projects/<encoded-cwd>/<session_id>.jsonl directly — no data replication, no network calls. Pure stdlib Python, zero pip dependencies. Persistent daemon HTTP server on port 7878 (auto-fallback 7879-7888 if busy; override with CONTEXT_TIMELINE_PORT env var). Watchdog auto-shutdown after 1h of inactivity. Disable browser auto-open with CONTEXT_TIMELINE_NO_BROWSER=1. Manual shutdown: python3 .claude/hooks/context-timeline.py --shutdown

Install Command

npx claude-code-templates@latest --hook monitoring/context-timeline

View on GitHub

claude-code — context-timeline

Scanned by NVIDIA SkillSpector

Automated static analysis (no LLM) · v2.12.0 · 2026-10-17

Partial analysis: some checks could not run on this component.

8 observations

scanner score 92/100

SkillSpector looks for patterns such as shell commands, credential access, network calls or unpinned dependencies. An observation is something worth reading in the files before you install, not a confirmed vulnerability. Components that run code locally, like hooks, MCP servers and security tooling, naturally show more of them.

Agent Config Directory Access ×2
Agent Snooping · context-timeline.py:75
Unbounded Resource Access ×2
Excessive Agency · context-timeline.py:147
Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: env, file_read, file_write, network, shell.
MCP Least Privilege · SKILL.md:1
subprocess module call
Dangerous Code Execution · context-timeline.py:1109
Internal Network Request
Server-Side Request Forgery · context-timeline.py:1173

Showing 7 of 8 observations. Learn more about NVIDIA SkillSpector.

Content

JSON

{

"description": "Real-time browser visualization of Claude Code's context window and subagent/too..."

"hooks": {

"SessionStart": [1 item]

"PreToolUse": [1 item]

"PostToolUse": [1 item]

"Stop": [1 item]

}

Context Timeline

Scanned by NVIDIA SkillSpector

Content

Stack Builder