Agent Token Saver
Route context through the smallest measured CLI or projection before loading heavy tools.
Install
$ agentstack add skill-supersynergy-agent-token-saver-agent-token-saver ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Links to your public security report.[](https://agentstack.voostack.com/security/report/skill-supersynergy-agent-token-saver-agent-token-saver) Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Agent Token Saver
Use this skill when a task can produce large logs, files, scraped pages, tool schemas, repeated research, or a broad skill catalog.
Default ladder
- Exact local search or deterministic projection.
- RTK for supported noisy shell commands.
- Load zero or one primary routed skill automatically, never the whole library.
- Use a budgeted code reader only when exact search misses.
- Enable graph, browser, or large-context tooling for one session only.
Keep approvals, safety rules, exact evidence and error lines intact. Never claim provider billing savings from a local character estimate.
Commands
agent-token-saver doctor --profile lean
python3 scripts/news_projection.py raw.jsonl --top 40 --format jsonl
python3 scripts/token_stack_matrix_benchmark.py --helpFor adaptive routing across a large skill collection, install the companion agent-token-saver-skill-router. Keep the index and ranking outside model context; load only the selected domain skill:
https://github.com/Supersynergy/agent-token-saver-skill-router
si route "" --max 1 --strict --json # automatic 0/1 decision
si find "" --json # manual discovery only
si resolve "" --json # exact path only
si index --refresh --json # after skill changesDo not auto-load the router skill itself or a second reserve skill. An explicit $SkillName remains the user override.
Done gate
- Same accepted result before and after optimization.
- Raw and optimized token estimate recorded.
- Latency recorded.
- Optional tools remain optional.
- Hook presence verified in live agent config, not inferred from files on disk.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: Supersynergy
- Source: Supersynergy/agent-token-saver
- License: MIT
- Homepage: https://github.com/Supersynergy/agent-token-saver#readme
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.