What is budget-term-sheet?

A cost-approval step for GPU jobs, where GPUs are processors commonly used for machine-learning training. It prepares a quote and requires explicit approval before a qualifying job is submitted.

Use it to quote budgets for training, PyTorch or Triton jobs, and model-weight downloads, then approve or reject the planned provider, GPU count, region, tier, and time limit.

It makes the expected compute cost visible before work starts and prevents unapproved or underfunded GPU submissions. It does not launch the job itself.

budget-term-sheet — Skill for Claude Code by Rockielab

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent

npx skills add Rockielab/rockie-claude --skill budget-term-sheet

Clone the repo

git clone --depth 1 https://github.com/Rockielab/rockie-claude

Made for: Claude Code.

Or install rockie-claude, the plugin that ships this one along with the rest of its 29 skills, 1 MCP server.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/rockielab/rockie-claude/budget-term-sheet/github.svg)](https://agentmods.dev/skills/rockielab/rockie-claude/budget-term-sheet)

Your own site

<a href="https://agentmods.dev/skills/rockielab/rockie-claude/budget-term-sheet"><img src="https://agentmods.dev/badge/skills/rockielab/rockie-claude/budget-term-sheet/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/rockielab/rockie-claude/budget-term-sheet"><img src="https://agentmods.dev/badge/skills/rockielab/rockie-claude/budget-term-sheet.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 68 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 911 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00068	$0.00911
Opus 5	$0.00034	$0.00456
Sonnet 5	$0.00014	$0.00182
Haiku 4.5	$0.00007	$0.00091

Measured 12d ago against content hash 0e64902489d4, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

budget-term-sheet scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.

The scan reads SKILL.md. This mod also ships 6 executable files (scripts/parse_approval.py, scripts/quote_term_sheet.py, scripts/render_term_sheet.py, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

Do not shell raw `curl POST /api/jobs/submit`.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

budget-term-sheet — 100% identical, 0 lines differ

project-harness/skills/budget-term-sheet/SKILL.md · 100 lines

How it starts

The opening of the file, as written. The whole thing — 100 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/budget-term-sheet

Render the user-facing cost term sheet that Rocky must show before a Rocky-originated POST $ROCKIELAB_API_URL/api/jobs/submit call. This skill is pre-submit UX and approval gating only. It does not replace platform-context's credit reservation, day caps, or hard budget enforcement.

When to invoke

Before /experiment submits a GPU / torch / triton / training / weight-download job.
When the user asks for a dry-run quote without launching compute.
When a user changes the approved budget and Rocky needs to re-render the quote.

Hard rules

Rocky-originated GPU submits go through this skill before runtime/submit.py.
Only explicit approval tokens count: approve, approved, or I approve.
A user-modified budget below estimate_cents is non-submittable in this repo.
Non-available market states never submit.
Keep the user-facing provider deidentified: Rockie GPU, plus SKU/count/region/tier.
Do not dispatch real compute here. Quote and approval only.

Procedure

Assemble a job spec with job_shape, gpu_type, gpu_count, region, tier, and either timeout_seconds or wallclock_minutes. Use region: "us" and tier: "spot" unless the user or workflow requires a different visible capacity slice.
Run scripts/quote_term_sheet.py to build the canonical JSON term-sheet object. Prefer live quote endpoints when credentials exist; use --market-json for tests and dry runs; use --allow-heuristic-dry-run only when the user explicitly wants a heuristic quote.
Render the markdown block with scripts/render_term_sheet.py. The block must show: total estimate, stage breakdown, recommended budget, wall-clock, Rockie GPU compute summary, availability/confidence, and the explicit Approve / Modify / Cancel prompt.
Wait for a user reply. Parse it with scripts/parse_approval.py.
Decision handling:
- approve: keep the current budget and mark the term sheet approved.
- modify: re-render with the new user_budget_cents and ask again.
- modify_then_approve: re-render with the new budget; only submit if the modified budget is still >= estimate_cents.
- cancel: abandon the submit.
- clarify: ask again; do not submit.
Hand the final approved JSON artifact to /experiment/runtime/submit.py through --term-sheet-json, passing matching --region and --tier. The submit helper refuses artifacts or CLI requests that omit or change the approved region/tier. Do not shell raw curl POST /api/jobs/submit.

Read the full file on GitHub · 100 lines

Files

What ships with it

10 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

12d ago First seen · 100 lines · 68 tokens per session scan A 0e64902489d4

Subscribe to this mod's changes

budget-term-sheet is a skill published in the GitHub repository Rockielab/rockie-claude (21 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 68 tokens to every session and 911 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Claude Code Data and AI Finance and fintech agent-harness ai-research anthropic autonomous-research claude

context-manager

Gestion avancée du contexte pour agents IA — fenêtre de contexte, compression, RAG dynamique, sliding window, prompt caching. Se déclenche avec "context window", "contexte agent", "token limit", "context management", "context overflow", "trop de contexte", "agent context", "context stuffing", "context compression".…

khalilbenaz/claude-skills-collection · 84 tokens

data-analyst-agent

Création d'agents d'analyse de données qui explorent, visualisent et expliquent les données. Se déclenche avec "data analyst agent", "agent analyse données", "agent data", "agent qui analyse", "automated analysis", "data exploration agent", "agent SQL", "agent pandas". Also triggers on "agent that analyzes data"…

khalilbenaz/claude-skills-collection · 81 tokens

openai-assistants-builder

Création d'assistants IA hébergés avec l'API OpenAI Assistants v2. File search avec vector stores, code interpreter, function calling, threads persistants et streaming. Se déclenche avec "OpenAI Assistants", "assistant API", "file search", "code interpreter", "thread", "run", "assistant OpenAI", "GPT assistant"…

khalilbenaz/claude-skills-collection · 116 tokens

security-hardener

Sécurisation d'agents IA contre injections, abus et fuites de données. Se déclenche avec "sécurité agent", "agent security", "prompt injection", "jailbreak", "agent abuse", "guardrails", "safe agent", "sécuriser mon agent", "agent en production sécurisé". Also triggers on "secure my agent", "prompt injection defense"…

khalilbenaz/claude-skills-collection · 86 tokens

anomaly-detection-builder

Conception de systèmes de détection d'anomalies (statistique, ML, deep learning) — arbre de décision, snippets copiables, anti-patterns, déploiement production. Se déclenche avec "anomaly detection", "détection d'anomalies", "outlier", "fraud detection", "anomalie". Also triggers on "outlier detection", "detect…

khalilbenaz/claude-skills-collection · 88 tokens

model-optimization-guide

Optimisation de modèles ML pour l'inférence (quantization, pruning, distillation, ONNX) — guide opérationnel avec snippets copiables, critères de choix et anti-patterns. Se déclenche avec \"quantization\", \"pruning\", \"distillation\", \"ONNX\", \"optimiser un modèle\", \"inférence rapide\". Se déclenche avec…

khalilbenaz/claude-skills-collection · 126 tokens

budget-term-sheet