Hugging Face's logo

Datasets:
LucioLiu
/
token-saver

Tasks:
Text Generation
Languages:
English
Chinese
Size:
n<1K
Tags:
agent-skill
agent-skills
claude-skill
claude-code
skill
agentskills
License:
Dataset card Files
xet
 Community
token-saver / SOP.md
LucioLiu's picture
LucioLiu
Initial release - mirrors github.com/LucioLiu/token-saver
31a7f1c verified
preview code
|
Raw
History Blame Contribute Delete
2.45 kB

English | 简体中文

token-saver · Manual SOP (runnable by a human, no AI required)

This is the skill's manual operation layer: SKILL.md compressed into a card a human can follow by hand — when you (a human) dispatch AI subagents manually, or want this discipline in a team runbook. The SKILL remains the rule source; this card compresses, it does not add rules.

  1. Sensitivity first: classify the data (PUBLIC / INTERNAL / CONFIDENTIAL / RESTRICTED), filter to models allowed to touch it, and only then discuss capability and cost (details: references/sensitivity.md).
  2. Judge task intensity (light execution / routine work / deep judgment / critical review) — don't go by the task's name; check five dimensions: uncertainty / cost of error / verifiability / context complexity / side effects. Any one clearly elevated → bump the tier up.
  3. Inventory on the spot: only consult the models / reasoning options the current platform actually exposes or permits; keep no long-lived model roster.
  4. Route by interface capability: use EXPLICIT when the interface lets you pin a model; PRESET when you can only pick predefined agents; INHERIT_ONLY when nothing is settable (label it honestly — never pretend a model was chosen). Log one ROUTE_DECISION record per dispatch (format: references/route-decision.md).
  5. Cascade has preconditions: only when output is machine-checkable and isolation / idempotency / rollback preconditions hold — run one tier lower first (never below light execution), on failed verification retry once at a higher tier, then stop and return failure evidence. Irreversible side-effect tasks (sending messages / touching databases / deploys / bulk deletes) and the critical-review tier must not cascade; under INHERIT_ONLY a rerun is a same-tier retry, never an "upgrade" (full terms: references/cascade.md).
  6. Verify after dispatch: confirm the actual model from host run records; if the host exposes usage data, include actual spend in the flow-back (session-local reference only); mark "unverified" when you can't confirm.
  7. Models bind to task intensity, never permanently to an agent's identity.

Maintained in sync with the SKILL; recheck this card whenever the SKILL version bumps. If they diverge, SKILL.md wins.