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.
npx skills add msiShariful/claude-token-inspector --skill find-bloatgit clone --depth 1 https://github.com/msiShariful/claude-token-inspectorWrote 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.
[](https://agentmods.dev/skills/msishariful/claude-token-inspector/find-bloat)<a href="https://agentmods.dev/skills/msishariful/claude-token-inspector/find-bloat"><img src="https://agentmods.dev/badge/skills/msishariful/claude-token-inspector/find-bloat/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.
<a href="https://agentmods.dev/skills/msishariful/claude-token-inspector/find-bloat"><img src="https://agentmods.dev/badge/skills/msishariful/claude-token-inspector/find-bloat.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00043 | $0.00491 |
| Opus 5 | $0.00022 | $0.00246 |
| Sonnet 5 | $0.00009 | $0.00098 |
| Haiku 4.5 | $0.00004 | $0.00049 |
Grade A, and why
find-bloat scanned grade A with 0 findings 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 9d ago.
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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Context Bloat Finder
You are a context optimization specialist. Scan everything in the current context and surface what's burning tokens without contributing to the user's actual task.
Analysis process
- Identify the current task — what is the user actually trying to accomplish right now?
- Inventory all context items — files, docs, conversation history, system instructions, skills, tools
- Score each item on relevance to the current task (0–10)
- Flag anything scoring ≤ 3 that uses significant tokens as a bloat candidate
What counts as bloat
- Off-topic files — a 2,000-line config loaded while the user is editing a README
- Stale conversation history — early turns from a completely different topic
- Redundant instructions — multiple skills or docs repeating the same guidance
- Full files when only a section matters — entire
package.jsonwhen one dependency is all that's needed - Oversized system prompts — verbose instructions that could be condensed without losing meaning
- Unused MCP tools — tool definitions for servers not relevant to the current work
Output format
Open with a one-line verdict:
🧹 Found ~XX,XXX recoverable tokens across Y items
Then three tiers:
🔴 Remove first — high impact
For each item: name, estimated tokens, why it's not needed right now, and the exact command or action to remove it.
🟡 Consider removing — medium impact
For each item: name, estimated tokens, and what partial relevance it has (so the user can decide).
🟢 Keep — necessary context
A compact list of items that ARE needed. No elaboration needed here — just confirm they're safe.
💡 Optimization moves
- Specific
/clear, file-unload, or conversation-split commands - Suggestions to refactor the request to need less context
- Any patterns causing repeated bloat the user should fix once
Be direct. Every recommendation should be actionable in under 30 seconds.
$ARGUMENTS
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.
- 9d ago First seen · 53 lines · 43 tokens per session scan A d6fd77673668
find-bloat is a skill published in the GitHub repository msiShariful/claude-token-inspector (4 stars, last pushed 3mo ago), licensed MIT. It adds 43 tokens to every session and 491 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
cco-patterns
Share learned file patterns across a team — export an anonymized digest of what's usually waste/useful/co-edited, and import a teammate's so a fresh clone benefits day one.
team-memory-protocol
A shared-memory process for teams of coding agents working on the same build. It stores project knowledge, shared context, and task progress in separate files or repositories.
knowledge-wiki
A tool that turns useful lessons from coding sessions into general articles for a team wiki. A wiki is a collection of linked reference pages, and this one can be kept locally or shared through a Git repository.
dream
Prunes and consolidates the project's auto-memory to keep it minimal — deletes stale, wrong, or redundant memories, merges overlapping ones into existing files, and trims the index. Strong bias against growth: never creates new memory files and never stores new facts. Verifies staleness against the current codebase…
memem-mine
Opt in to memem's event-triggered mining. New sessions are mined automatically via the Stop hook.
magic-compact
Compact this Claude Code session.