token-optimizer

Token Optimizer is a local (external Python) tool/plugin that audits and reduces “ghost tokens” in Claude Code / OpenClaw sessions by measuring context degradation, detecting wasted/duplicate content, tracking compaction loss, and providing checkpoint/restore and archival tooling for tool results.

Evaluated Mar 30, 2026 (22d ago)
Repo ↗ DevTools ai-ml devtools context-window token-optimization claude-code-plugin openclaw-plugin
⚙ Agent Friendliness
46
/ 100
Can an agent use this?
🔒 Security
29
/ 100
Is it safe for agents?
⚡ Reliability
35
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
0
Documentation
55
Error Messages
0
Auth Simplicity
90
Rate Limits
20

🔒 Security

TLS Enforcement
50
Auth Strength
20
Scope Granularity
0
Dep. Hygiene
40
Secret Handling
40

README claims execution as an external Python process and provides features like caching, reading, and archiving tool results to disk, but it does not describe threat model, data handling/retention, TLS usage specifics, or secret redaction practices. curl|bash installer increases supply-chain risk if not pinned/verified. Authentication/scoping for any network API is not applicable/unspecified.

⚡ Reliability

Uptime/SLA
0
Version Stability
45
Breaking Changes
40
Error Recovery
55
AF Security Reliability

Best When

You frequently run long Claude Code / OpenClaw sessions with heavy custom skills/MCP/tooling overhead and you experience fast context degradation or frequent compactions.

Avoid When

You cannot allow local filesystem access/read-cache/archival artifacts, or you need strict governance on code execution (e.g., disallowing bash installers).

Use Cases

  • Diagnose context/window overhead (“ghost tokens”) in Claude Code sessions
  • Improve prompt efficiency by blocking redundant file reads and ignoring files via .contextignore
  • Measure session quality drift over time (7-signal quality score and efficiency grading)
  • Recover from compaction by restoring checkpoints and expanding archived tool results
  • Analyze and trim large JSONL session artifacts and detect duplicates (tooling utilities)
  • Restructure CLAUDE.md to improve attention placement (attention-optimize)

Not For

  • Producing a hosted, centrally-managed analytics service (it appears to run locally per platform)
  • Security-critical workflows where you must guarantee zero data retention on remote services (the README does not provide such guarantees)
  • Environments where you cannot execute the provided installer script (curl | bash)

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
No
SDK
No
Webhooks
No

Authentication

OAuth: No Scopes: No

No authentication mechanism is described for Token Optimizer itself; it appears to operate as a plugin/script alongside Claude Code/OpenClaw and external Python tooling.

Pricing

Free tier: No
Requires CC: No

Pricing for the package is not provided in the provided content.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Installer uses a curl|bash approach; automated agents should prefer pinned, reviewed installation steps when possible.
  • Behavior depends on environment variables/config (e.g., read-cache modes); ensure correct settings per environment.
  • Local caching/archival implies filesystem artifacts; agents should clean up or manage storage if running many sessions.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for token-optimizer.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered