tldr

Featured

TLDR code analysis — call graphs, semantic search, impact, dataflow, for far fewer tokens than reading the files raw. Triggers "who calls X", "what affects X", "blast radius", before large file reads or refactors.

Data & Documents 44 stars 3 forks Updated today MIT

Install

View on GitHub

Quality Score: 90/100

Stars 20%
55
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
100
Issue Health 10%
50
License 10%
100
Description 5%
100

Skill Content

# TLDR Code Analysis ## Standalone Codex fallback cc-settings does not install the TLDR MCP into standalone Codex. Ignore the Claude `context`, `allowed-tools`, and `requires` frontmatter in that host and keep the task read-only with native tools. Use `rg --files` to map the tree, `rg -n '<symbol|pattern>'` for exact references, direct import searches such as `rg -n 'from .*<module>|require\(.*<module>'`, and caller searches for the symbol followed by focused file reads. Use `git diff --name-only` plus test-name and import searches for change impact. Do not invoke or claim to have invoked a TLDR MCP in standalone Codex. Report the native searches actually run and their limitations. The remaining workflow is for Claude hosts with the configured TLDR MCP. Token-efficient codebase analysis behind the `tldr` MCP server. It returns the symbols, edges, and slices you asked for instead of whole file bodies, so a question that would cost several full reads costs one small structured answer. > No measured savings figure is published here on purpose. cc-settings carried a "~95% fewer tokens" claim for months with no benchmark behind it anywhere in the repo — the kind of number `AGENTS.md` now forbids (*No savings against a run that never happened*). To get a real figure, answer the same question both ways and compare the token counts your own session reports. The engine is provisioned by cc-settings. The tool names below are the stable contract; only the engine behind them change...

Details

Author
darkroomengineering
Repository
darkroomengineering/cc-settings
Created
8 months ago
Last Updated
today
Language
TypeScript
License
MIT

Integrates with

Bundled in these plugins

Similar Skills

Semantically similar based on skill content — not just same category

AI & Automation Listed

codeindex

Answer code questions structurally instead of reading whole files. Use when you need to find where something is defined, read one function, see what calls a symbol, trace imports, judge what a change breaks, or orient yourself in an unfamiliar file or repository. Reach for this before Read, Grep or Glob on a code question — and equally before `grep`, `rg`, `cat`, `head`, `sed -n` or `find` in a shell, which is where the same question goes when a harness or permission mode routes file work through Bash. A grep finds a string; it cannot tell you what calls what. Backed by the codeindex MCP server (tree-sitter, 40+ languages).

3 Updated today
munhq
AI & Automation Listed

tokenscope

Judgment-layer review of Claude Code token usage. Use when the user asks to audit token costs, review Claude Code spending, check context waste, or interpret a tokenscope report. Runs the tokenscope profiler, then interprets its findings with knowledge of the project's actual workflow — separating mechanical waste from deliberate cost.

7 Updated 1 months ago
AviVAvi
AI & Automation Listed

token-efficiency

Token optimization for Claude Code CLI sessions: RTK (Rust Token Killer) command-output compression, efficient file and command strategies, role-based model routing for subagents, Source-of-Truth keyword framework, hybrid SQLite agent memory with checkpoints, and concise output. Use when installing or verifying RTK or its hook, when a dev command's output is large (test runs, builds, installs, long logs), when spawning coding subagents, when writing agent logs or querying agent memory, when seeding a project's SoT keyword index, or when folding agent memory at a checkpoint. Also use when the user mentions 'rtk', 'token efficiency', 'token savings', 'context bloat', 'model routing', 'SoT keywords', or 'checkpoint agent memory', or asks for shorter or less verbose output, or to stop narrating or dumping whole files.

1 Updated 1 months ago
M4NUSH7