Stats
Actions
Tags
From context_forge
This skill should be used when the user says "estimate token usage", "how many tokens am I using", "token cost of my setup", "audit token budget", "what's eating my context", "token usage breakdown", or "how expensive is my session".
How this skill is triggered — by the user, by Claude, or both
Slash command
/context_forge:estimate-tokensThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Audit and report approximate token usage across all context sources: claude.md, skills, memory files, plugins, and conversation history.
Audit and report approximate token usage across all context sources: claude.md, skills, memory files, plugins, and conversation history.
1 word ≈ 1.3 tokens
1 line of code ≈ 10-15 tokens
1KB of text ≈ 200-300 tokens
wc -w ~/.claude/CLAUDE.md 2>/dev/null || echo "No global CLAUDE.md"
wc -w ./CLAUDE.md 2>/dev/null || echo "No project CLAUDE.md"
Report: CLAUDE.md: ~[words] words ≈ [words * 1.3] tokens
find ~/.claude/plugins -name "SKILL.md" -exec wc -w {} \; 2>/dev/null | sort -rn | head -20
Sum the word counts. Multiply by 1.3 for token estimate.
wc -w ~/.claude/memory/*.md 2>/dev/null
find . -path ".remember/*.md" -exec wc -w {} \; 2>/dev/null
ls ~/.claude/plugins/ 2>/dev/null | wc -l
Each active plugin = at minimum plugin.json metadata overhead.
Output formatted breakdown:
TOKEN USAGE AUDIT
=================
claude.md (global): ~[N] tokens
claude.md (project): ~[N] tokens
Skills loaded: ~[N] tokens ([X] skills)
Memory files: ~[N] tokens
─────────────────────────────────
TOTAL CONSTANT COST: ~[N] tokens per response
Top token consumers:
1. [skill-name]: ~[N] tokens
2. [skill-name]: ~[N] tokens
3. [memory-file]: ~[N] tokens
Recommendation: [action to reduce]
Based on audit results, recommend in order:
Emit structured output as LTX rows when reporting per-source token estimates.
@v1:source|words|tokens|status
| Field | Description |
|---|---|
source | File or category (e.g. ~/.claude/CLAUDE.md, skills) |
words | Raw word count |
tokens | Estimated tokens (words * 1.3, rounded) |
status | ok, warn (approaching limit), critical (over limit) |
Example:
@v1:source|words|tokens|status
~/.claude/CLAUDE.md|850|1105|critical
./CLAUDE.md|320|416|ok
skills|2400|3120|warn
references/token-benchmarks.md — Token cost benchmarks for common setupsnpx claudepluginhub kyuna0312/context_forge --plugin context_forgeSearches MemPalace before answering questions about past work, people, projects, or prior decisions. Returns verbatim stored content instead of guessing from model memory.
Guides Payload CMS config (payload.config.ts), collections, fields, hooks, access control, APIs. Debugs validation errors, security, relationships, queries, transactions, hook behavior.
Implements vector databases with Pinecone, Weaviate, Qdrant, Milvus, pgvector for semantic search, RAG, recommendations, and similarity systems. Optimizes embeddings, indexing, and hybrid search.