token-budget-guard SKILL.md skillA
token-budget-guard from ducdg88/ducpt-skills is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add ducdg88/ducpt-skills/token-budget-guard@v1.0.0Pins v1.0.0 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- ducdg88/ducpt-skills
- Path
skills/token-budget-guard/SKILL.md- Versions recorded
- 1
- Latest version
- v1.0.0, committed Sep 27, 2026
- Size
- 4,147 bytes
- Repository stars
- 8
- License
- MIT
Read local Claude Code or Claude API usage logs and report token counts, prompt-cache hit rate and estimated cost by model, then warn before a session goes over a spending budget. Runs fully offline on files already on disk; never calls a paid API and never guesses prices. Use when the user says "tiết kiệm token", "đang đốt credit", "token budget", "chi phí AI", "cảnh báo trước khi vượt ngân sách", "cache hit là gì", or wants to know why an agent session cost more than expected before running…
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# Token Budget Guard Agent tool bills often surprise people: credit gets deducted with no clear reason, or a long multi-agent run costs far more than expected. This skill answers "where did it go" and stops the next run before it repeats, using only logs already on the machine. ## Rule: never guess a price Token counts come straight from the log and are always exact. Cost is only computed when the user supplies `prices.json` with numbers they read themselves from the provider's current pricing page. A price typed from memory here would silently give a wrong budget check, so this skill refuses to estimate cost without it (see `data/prices.example.json` for the shape; those numbers are placeholders, not real prices). ## Step 1. Find the log Claude Code keeps a JSONL transcript per session (ask the user where theirs is, or look under the tool's own session/history folder). Any JSONL or JSON file works as long as some records contain a `usage` object with `*_tokens` fields; the script finds these wherever they are nested and does not need an exact schema match. ## Step 2. Report ```bash python scripts/token_budget.py report session.jsonl …
Read the whole file at its exact version.
How to install
mdr add ducdg88/ducpt-skills/token-budget-guard@sha256:bc62e1ead2839b2bThe line at the top of this page pins the label v1.0.0, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
[](https://markdownregistry.com/a/art_xifhjrchvv423wtz)
4 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 1 version of this file, committed on Sep 27, 2026.
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (4147 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
ducdg88/ducpt-skills · 8 stars · license MIT · pushed 2026-10-23 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_xifhjrchvv423wtz GET https://markdownregistry.com/api/v1/resolve?ref=ducdg88/ducpt-skills/token-budget-guard GET https://markdownregistry.com/api/v1/blob/bc62e1ead2839b2b81912d53b4df33795c7a04bfd08d92f18a485c97b478d1e8
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.