fix(context-timeline): detect Opus 1M context window dynamically

[mario-hernandez]

Problem

CONTEXT_LIMIT is hardcoded to 200_000 tokens in context-timeline.py. Users on Claude Opus 4.7 with the 1M context window ("model": "opus[1m]" in ~/.claude/settings.json) see the dashboard report 80–90% usage when actual usage is 16–18%, producing constant false alarms and making the meter unusable on long sessions.

Fix

Replace the constant with _detect_context_limit(cwd) which resolves the limit in this order:

1. CONTEXT_TIMELINE_LIMIT env var — explicit override (e.g. export CONTEXT_TIMELINE_LIMIT=500000).
2. [1m] / [1000k] / [200k] suffix in the model field of <cwd>/.claude/settings.json or ~/.claude/settings.json.
3. 200_000 fallback — current behaviour, untouched for users without [1m].

The limit is resolved once per _Builder instance, not per update tick — zero per-event filesystem cost.

Backwards compatibility

• No new dependencies (still stdlib only).
• Default behaviour unchanged for users without [1m] in settings.
• No new hooks, no breaking API changes.

Verification

Tested live on a real Opus 1M session of ~178K tokens:

	Before	After
Header reading	178,581 / 200,000 (89%)	178,581 / 1,000,000 (17.9%)
Bar colour	Red / warning	Cyan / normal
Sidebar meter	Saturated	Accurate

Resolution order also unit-tested:

• env var override → int(env)
• project .claude/settings.json with [1m] → 1_000_000
• global ~/.claude/settings.json with [1m] → 1_000_000
• no settings → 200_000

Files changed

cli-tool/components/hooks/monitoring/context-timeline.py (+44, −8)

---

Summary by cubic

Fixes incorrect context-usage reporting by dynamically detecting the model’s context window, so Opus 1M sessions show accurate percentages instead of false alarms. Replaces the hardcoded 200k limit with a detected value.

• Bug Fixes
  • Area: components (cli-tool/components/hooks/monitoring/context-timeline.py).
  • Detects limit once per session via _detect_context_limit(cwd): env var CONTEXT_TIMELINE_LIMIT → [1m]/[1000k]/[200k] in .claude/settings.json → 200_000 fallback.
  • No new components; no docs/components.json regeneration needed.
  • New optional env var: CONTEXT_TIMELINE_LIMIT (int). No secrets.

^{Written for commit 6c3750a. Summary will update on new commits. Review in cubic}

[vercel]

@mario-hernandez is attempting to deploy a commit to the Daniel Avila's projects Team on Vercel.

A member of the Team first needs to authorize it.

[github-actions]

👋 Thanks for contributing, @mario-hernandez!

This PR touches cli-tool/components/** and has been marked review-pending.

What happens next

1. 🤖 Automated security audit runs and posts results on this PR.
2. 👀 Maintainer review — a human reviewer validates the component with the component-reviewer agent (format, naming, security, clarity).
3. ✅ Merge — once approved, your PR is merged to main.
4. 📦 Catalog regeneration — the component catalog is rebuilt automatically.
5. 🚀 Live on aitmpl.com — your component appears on the website after deploy.

While you wait

• Check the Security Audit comment below for any issues to fix.
• Make sure your component follows the contribution guide.

This is an automated message. No action is required from you right now — a maintainer will review soon.

[cubic-dev-ai]

No issues found across 1 file