BTreeMap

17 mods across 1 repository, 2 stars between them.

btm-skills

01

BTreeMap/SKILLs

Plugin Claude Code

Reusable, project-agnostic agent skills from the BTreeMap/SKILLs library.

2 yesterday A tokens not measured original MIT

btm-skills

02

BTreeMap/SKILLs

Plugin Claude Code

A library of reusable, project-agnostic agent skills: each one a SKILL.md procedure an LLM coding agent loads and follows.

2 yesterday A tokens not measured original MIT

SKILLs AGENTS.md

03

BTreeMap/SKILLs

Instructions file CodexOpenCode

Instructions for BTreeMap/SKILLs, covering agents.md, repository shape, automated gate, frontmatter protocol and authoring and editing skills.

2 yesterday A 5,924 tokens original MIT

aesthete

04

BTreeMap/SKILLs

Skill Claude CodeCodex

Designs, builds, reviews, and reworks user interfaces with an HCI researcher's rigor, an art director's eye, and a type theorist's structural discipline: ingests a supplied design document and color palette under a strict precedence ladder, enforces hierarchy, typography, color, spacing, motion, interaction-state, and…

2 yesterday A 199 tokens original MIT

author-skill

05

BTreeMap/SKILLs

Skill Claude CodeCodex

Distills a completed task history, workflow, or procedure into a reproducible SKILL.md that a fresh agent can execute without external memory, and reviews existing skills for conformance to the Agent Skills standard. Use when the user asks to create, refactor, review, or distill a skill, or when a repeated procedure…

2 yesterday A 73 tokens original MIT

caveman

06

BTreeMap/SKILLs

Skill Claude CodeCodex

Compresses language model output into token-economical caveman phrasing while keeping full technical accuracy: articles, filler, pleasantries, and hedging die; code, numbers, negations, and error strings stay exact. Supports intensity levels lite, full (default), ultra, and wenyan (classical Chinese) variants, plus…

2 yesterday A 165 tokens original MIT

fact-check

07

BTreeMap/SKILLs

Skill Claude CodeCodex

Verifies factual claims in documents by decomposing them into atomic claims, retrieving evidence from authoritative sources, and proposing corrections with calibrated verdicts, verbatim quotes, URLs, and access dates. Edits apply only after explicit user approval. Auto-detects environment capabilities: parallel…

2 yesterday A 147 tokens original MIT

git-commit

08

BTreeMap/SKILLs

Skill Claude CodeCodex

Drafts and reviews git commit messages that follow Conventional Commits, resolving scopes from repository history while minimizing output tokens. Supports effort levels: lite (subject line only, no history scan), full (default: scoped subject plus wrapped body and footer), and ultra (adds an atomicity and…

2 yesterday A 114 tokens original MIT

humanize

09

BTreeMap/SKILLs

Skill Claude CodeCodex

Rewrites AI-sounding text so it reads like its writer without changing what it says: detects 35 patterns from Wikipedia's "Signs of AI writing" (inflated claims, sales language, vague sources, overused AI words, formulaic rhetoric, chatbot artifacts, dash and formatting tells), preserves every claim, invents no facts…

2 yesterday A 156 tokens original MIT

lit-review

10

BTreeMap/SKILLs

Skill Claude CodeCodex

Runs a literature review as a staged, evidence-tracked pipeline: fixes the question and inclusion criteria before any search, searches OpenAlex, arXiv, and Crossref through a bundled keyless script that logs every query and deduplicates by DOI, arXiv id, and title, screens candidates in two passes with recorded…

2 yesterday A 202 tokens original MIT

pl-theorist

11

BTreeMap/SKILLs

Skill Claude CodeCodex

Applies a programming-languages theorist's discipline across the software lifecycle through verbs - design, build, refactor (default), review, audit, test, teach, help - producing immutable data, algebraic data types, smart constructors, total domain models, explicit effects, complexity-conscious data structures (hash…

2 yesterday A 205 tokens original MIT

ponder

12

BTreeMap/SKILLs

Skill Claude CodeCodex

Answers open questions, shower thought to academic grade, with uniform investigative rigor: the lead agent runs the first round itself and stops early when a question settles there, decomposes harder questions into retrievable leaves tracked in a script-owned ledger, fans a few orthogonal workers over bundles of…

2 yesterday A 149 tokens original MIT

ponytail

13

BTreeMap/SKILLs

Skill Claude CodeCodex

Forces the laziest solution that actually works - the simplest, shortest, most minimal. Channels a senior developer who questions whether the task needs to exist at all (YAGNI) and reaches for the standard library before custom code, native platform features before dependencies, one line before fifty. Levels: lite…

2 yesterday A 226 tokens original MIT

read-pdf

14

BTreeMap/SKILLs

Skill Claude CodeCodex

Reads PDFs with pypdf to extract text and metadata and answer questions with page-cited evidence; accepts a local file or an http(s) URL, fetching URLs itself into a temp cache; never creates, modifies, merges, splits, renders, OCRs, or otherwise writes PDF files. Use when the user asks to inspect, summarize, search…

2 yesterday A 95 tokens original MIT

reframe

15

BTreeMap/SKILLs

Skill Claude CodeCodex

Produces a testable strategic direction judgment: sharp thesis, constraint diagnosis, clean target, kill list, Conservative-vs-Clean-vs-Staged options, proof point, falsifier, and concrete payoff ledger. Use when the user asks to challenge a conservative design, think bigger, define a greenfield target, escape…

2 yesterday A 110 tokens original MIT

setup-env

16

BTreeMap/SKILLs

Skill Claude CodeCodex

Provisions an isolated, per-project development environment entirely in userspace: no sudo, no docker, no global installs, assuming only uv on PATH. A typed target algebra (family[:flavor][@version] tags such as [email protected], kotlin:android, go:cgo, rust, typescript, haskell) selects exactly the toolchains a project…

2 yesterday A 221 tokens original MIT

thematic-analysis

17

BTreeMap/SKILLs

Skill Claude CodeCodex

Develops themes from qualitative text under one named methodological school: reflexive, codebook, template, framework matrix, rapid, or hybrid inductive/deductive. Records the choice with its reason, builds a bounded codebook, codes every unit, runs a pre-declared agreement check where the school calls for one, and…

2 yesterday A 167 tokens original MIT