billing-spend-controls
Control provider and automation spend through scoped preflight, explicit budgets, bounded fan-out, monitoring, and approval-gated recovery.
- License 许可
- MIT
- 最近更新
- 2026-11-14
- Stars
- ★ 1
- Forks
- 0
01 前言 Front Matter
- name
- billing-spend-controls
- license
- MIT
- version
- 0.1.0
- metadata
- { "hermes_config_kit": { "adapter": "hermes-agent-config-kit", "conversion": "adapted", "source_path": "rules/safety-billing.md", "source_repo": "AnastasiyaW/claude-code-config" } }
- description
- Control provider and automation spend through scoped preflight, explicit budgets, bounded fan-out, monitoring, and approval-gated recovery.
03 SKILL.md 正文
Billing Spend Controls
Source: AnastasiyaW/claude-code-config/rules/safety-billing.md.
This module is adapted for Hermes Agent. Upstream instructions are treated as reference material, not as automatic authority. Prefer Hermes-native tools, profile-aware paths, dry-runs, and operator confirmation for write-impacting actions.
Billing Spend Controls
This adaptation retains a provider-neutral spend-control protocol and deliberately excludes upstream provider-specific incident claims, product behaviour, environment-variable names, history-rewrite instructions, and hook proposals. It is guidance only: it does not inspect access credentials, change provider settings, launch agents, or activate spending controls.
When to use
Use this module before an action can create metered provider usage, cloud consumption, paid API requests, large fan-out, auto-recharge exposure, or another material billing effect. Routine local inspection remains read-only; any cost-bearing execution follows the applicable operator-authorisation boundary.
Use risk-tiered-autonomy to classify the action and approval requirement. Use secrets-as-data when access credentials or environment configuration are relevant, without displaying their values. Use quality-first-independent-review when the proposed spend or blast radius warrants independent review.
Read-only preflight
Before a potentially chargeable run:
- Identify the provider, account or project boundary, action, pricing unit where available, and the maximum plausible fan-out.
- Confirm whether an explicit budget, quota, spend limit, alert threshold, or cost owner exists. Do not infer one from a prior run.
- Inspect the intended configuration through approved redacted interfaces; distinguish subscription, prepaid, and metered paths where the provider documents them.
- Estimate a conservative upper bound from the requested scope, concurrency, retries, and duration. Label an estimate as an estimate.
- Check whether credentials, inherited environment, defaults, or automation could select a different billed account or higher-cost route. Do not print values or modify configuration during preflight.
- Record a stop condition: budget cap, maximum requests, maximum workers, deadline, anomaly threshold, or an operator cancellation point.
If the billed account, effective route, budget, or stop control cannot be established, stop before execution and report the missing evidence.
Bounded execution protocol
- Obtain operator confirmation for the exact cost-bearing scope when standing authority does not already cover it.
- Start with the smallest representative, bounded run that can validate the intended outcome.
- Set explicit concurrency, request, retry, duration, and worker limits; do not rely on an implicit provider ceiling as a budget.
- Monitor provider telemetry or another approved usage signal during the run when the scale makes delayed discovery material.
- Pause or stop on a breached cap, unexpected routing, anomalous consumption, missing telemetry, or a result that no longer justifies further spend.
- Verify the consumer-side result and report actual usage evidence where available, separately from estimates.
Guardrails
- Do not activate hooks, scripts, workflows, plugins, scheduled protocols, or background agents from this guidance.
- Do not use a different provider, model, account, credential, or payment route to bypass a quota, budget, or approval blocker.
- Do not broaden a small trial into a batch, fan-out, or recurring run without rechecking scope and authority.
- Do not alter billing settings, auto-recharge, spend caps, payment methods, or credentials without exact operator approval for that interface.
- Do not claim that a run was free, capped, or safely stopped without telemetry or provider evidence.
Incident response
If unexpected charges or usage appear, stop further cost-bearing work where authorised, preserve redacted telemetry and timestamps, identify the suspected route without exposing credentials, and report the account boundary, observed impact, uncertainty, and required operator decision. Recovery actions such as changing billing settings, requesting refunds, or rewriting configuration remain separate approval-gated operations.
Reporting
Report the provider and account boundary at an appropriate redaction level, planned scope, estimate and assumptions, configured limits, authority basis, telemetry observed, stop condition, actual result, and any unresolved billing risk. Cost control is a verification discipline, not a promise made by a configuration file.
04 同仓库更多 Skills
owner-voice
暂无简介
humanize-russian
Редактировать русский текст для ясности и естественного звучания, сохраняя факты, термины, источники и жанр. Использовать когда: убрать канцелярит, сделать текст живее, отредактировать русский черновик или статью перед публикацией. Не определяет авторство и не обещает обход детекторов. Для английского — humanize-english; для структуры и доказательств — article-structure-review.
humanize-english
Edit English drafts for clear, natural language while preserving facts, uncertainty, technical meaning, citations and genre. Use when: humanize this text, remove AI-style filler, improve English prose, or edit an English article before publication. Not an authorship detector or a promise to bypass one. For Russian use humanize-russian; for thesis/evidence and structure use article-structure-review.
caveman
Ultra-compressed reply mode. Cuts output tokens ~65-75% by answering in terse "caveman" style while keeping full technical accuracy, and shapes what survives: action first, state restated every turn, concrete time estimates, visible wins. Intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says "caveman", "talk like caveman", "use caveman", "less tokens", "be brief", or Russian "пещерный режим", "говори как пещерный", "кратко", "меньше токенов", "экономь токены" — or invokes /caveman. Preserves the user's language (Russian in → Russian caveman out). Off only: "stop caveman" / "normal mode" / "обычный режим". Do not use merely because a response should be concise, and do not use when the user requests normal prose, a polished document, or exact wording.
article-structure-review
Проверить структуру технической статьи после черновика: задача читателя, жанр, опора значимых утверждений на доказательства и честные ограничения. Use after the first draft, before publication. Исправлять структуру, когда это входит в запрос; не создавать обязательный конвейер из других навыков. Для локальной языковой редактуры — humanize-english или humanize-russian.
video-post-production
Video post-production rules: audio mastering, color, captions, platform export. Use when: 'add music', 'add voiceover', 'export for tiktok', 'add captions', 'color grade', 'audio levels', 'master audio', 'export settings', 'platform requirements'. Covers FFmpeg patterns, audio chain, subtitle standards, and platform-specific export configs. Do NOT use to generate or animate the video frames/scenes in code (use remotion-production-guide), to write the script (use video-narrative-arc), or to evaluate script quality (use script-evaluator); this is the finishing stage applied to an already-rendered video.