{ "markdown": "

\n \"Julius\n

\n\n# Julius Skills\n\nEight personal agent skills for now: Caveman base, Interface Kit, Grill Me, Loop Factory, Junior to Senior, Deslopify, Context Canary, and The Last 20%.\n\nThis repo is shaped by eight things:\n\n- **Caveman** - 70k-star token compression without technical loss. Small mouth, big brain.\n- **Interface Kit** - accessible, performant interfaces with strong aesthetic direction, not generic AI slop.\n- **Grill Me** - calibrated pressure before hard critique, so challenge matches user knowledge and comfort.\n- **Loop Factory** - spec-driven agent loop where tasks move through inbox → active → archive with a real review gate.\n- **Junior to Senior** - adversarial senior review that treats agent output as junior work and upgrades it with codebase + web research.\n- **Deslopify** - mechanical scan-and-rewrite loop that erases AI-writing tells from any text and lands it in the right register.\n- **Context Canary** - per-turn canary signal that makes silent context degradation visible, plus a recovery protocol when it trips.\n- **The Last 20%** - finds and finishes the experiential layer agents skip: what the output reads like, which pages exist, first-run, defaults, the golden path.\n\nPoint is control. Agents should be terse when talking, precise when building interfaces, calibrated when challenging plans, and disciplined when running build loops.\n\n## Quickstart\n\nInstall with `skills.sh`:\n\n```bash\nnpx skills@latest add JuliusBrussee/skills\n```\n\nPick the skills you want for Claude Code, Codex, Gemini, Cursor, Windsurf, Cline, Copilot, or any agent supported by the installer.\n\nFor the canonical Caveman installer, use:\n\n```bash\ncurl -fsSL https://raw.githubusercontent.com/JuliusBrussee/caveman/main/install.sh | bash\n```\n\n## Why These Skills Exist\n\nMost agent output fails in three boring ways: too many words, UI that looks like every other generated demo, or critique that starts too hard before understanding user context.\n\nThis repo keeps three fixes close:\n\n1. **Speak less, say more.** Caveman cuts output tokens while preserving exact commands, code, errors, and technical meaning.\n2. **Build interfaces with taste.** Interface Kit starts from accessibility, performance, typography, spatial rhythm, color roles, and concrete interaction states.\n3. **Challenge at the right altitude.** Grill Me assesses knowledge and desired pressure first, then asks one question at a time.\n\n## Skills\n\n### `caveman`\n\nUltra-compressed communication mode. Cuts token usage by dropping filler, hedging, pleasantries, and excess grammar while keeping technical accuracy intact.\n\nUse when you want:\n\n- shorter agent replies\n- less token waste\n- exact code, command, error, and API preservation\n- persistent terse style until user exits with \"normal mode\"\n\n### `interface-kit`\n\nImplementation guide for high-quality UI. Synthesizes accessibility, performance, typography, layout, color systems, motion, interaction states, and component craft.\n\nUse when building:\n\n- frontend components\n- landing pages\n- dashboards\n- design systems\n- polished app flows\n- accessibility fixes\n- UI review passes\n\n### `grill-me`\n\nCalibrated interview skill for stress-testing plans, designs, and decisions. It first asks how much the user knows and how hard they want the pressure, then ramps from clarifying questions to failure-mode critique.\n\nUse when you want:\n\n- plan critique without getting overwhelmed\n- one question at a time\n- recommended answers with each question\n- pressure matched to beginner, working, or expert knowledge\n- softer or harder grilling on command\n\n### `loop-factory`\n\nSpec-driven agent loop. Coding tasks live as markdown specs that move through `inbox → active → archive`, get implemented by Claude Code or Codex, and must pass a review gate before they count as done. State is just which folder a spec is in. The governing rule: automate implementation and verification, not product decisions.\n\nUse when you want:\n\n- repeatable, reviewable agent work instead of one-off prompting\n- visible task state (inbox / active / archive) with no dashboard\n- generated implementation, review, and backprop prompts for either agent\n- a hard review gate before anything is marked done\n- to install or scaffold the `loop-factory` CLI into a project\n\nPairs with the [Loop-Factory](https://github.com/JuliusBrussee/Loop-Factory) repo, which ships the CLI and native Claude/Codex adapters.\n\n### `junior-to-senior`\n\nAdversarial review skill for agent-generated plans. Treats the current output as the work of a junior, then constructs a senior reviewer grounded in codebase research and web research of current best practices. Diagnoses altitude failures — plans that are foggy on the hard parts or tunneled into details with no product vision — and rewrites them into a scoped, state-of-the-art version with evidence behind every finding.\n\nUse when you want:\n\n- a staff-engineer-grade review of a plan before committing to it\n- plans that commit on interfaces, versions, and failure modes instead of hand-waving\n- best practices refreshed past the model's training cutoff via live web research\n- a clear delta between the original plan and the upgraded one\n- product decisions surfaced as open questions instead of silently invented\n\n### `deslopify`\n\nDe-slop pass for any text. Scans with a regex catalog of AI-writing tells — negative parallelism (\"not X but Y\"), puffery vocabulary, rule-of-three, false ranges, em-dash abuse, uniform cadence, hedged both-sidesing — then rewrites at the level of meaning and re-scans its own output until clean. Built as a loop because the worst tells are emergent generative habits that a rewriting model reintroduces in paraphrase.\n\nUse when you want:\n\n- AI-drafted text that reads like a person wrote it\n- a diagnosis table of which tells were found before the rewrite\n- register-aware output: academic article, tweet, reddit post, LinkedIn, email, blog, docs, marketing\n- no overcorrection — no fake typos, forced slang, or invented specifics\n\n### `context-canary`\n\nEarly-warning system for long agent sessions. Installs a byte-stable first-line signal — the user's name, a turn counter, and an honest context self-check — so the moment the agent's hold on its instructions degrades (attention drift, compaction, truncation), the signal visibly dies. Comes with a trip protocol: checkpoint state to a file, re-anchor on project instructions, reset deliberately.\n\nUse when you want:\n\n- to know *when* a long session starts rotting instead of finding out from bad output\n- a zero-infrastructure health check that runs every single turn\n- compaction events surfaced the moment they happen\n- a disciplined recovery path (checkpoint → re-anchor → fresh session) instead of limping on\n\nGrounded in context-rot research (Chroma), lost-in-the-middle (Liu et al.), and instruction-drift findings — sources linked in the skill's references.\n\n### `last-20-percent`\n\nFinds and finishes the last 20% of a built solution — the experiential layer agents skip. Agents decompose the noun (\"wiki\" → ingestion, search, LLM) and the experience of using the thing never appears in that decomposition, so it never becomes a task. This skill decomposes the *scene* instead: one concrete magic moment, walked step by step, with the residue specced at plumbing fidelity and hand-crafted golden artifacts set as the quality bar before any generator gets built.\n\nUse when you want:\n\n- builds that end with a product, not a technically-complete demo\n- the experiential work (content, IA, first-run, defaults, microcopy) specced as concretely as the plumbing\n- golden artifacts written by hand before the machinery that generates them\n- a final walk-through as the end user before anything is called done\n- an audit mode for existing solutions that work but feel flat\n\n## Interface Kit Standard\n\nIf a repo has `DESIGN.md`, it wins. Otherwise UI work should still have a point of view:\n\n- Accessibility first: contrast, keyboard navigation, focus states, semantics.\n- Performance before decoration: stable layout, lazy assets, transform-only motion.\n- Typography matters: readable type scale, sane line length, tabular numbers for data.\n- Layout uses stable spatial rules: 4/8px spacing, responsive constraints, no overlap.\n- Components have complete states: hover, focus, active, disabled, loading, empty, error.\n- Avoid generic defaults: no faceless hero, no purple-blue gradient template, no stock-looking polish.\n\n## Caveman Ecosystem\n\nThis repo sits next to broader Julius agent stack:\n\n| Repo | What |\n|---|---|\n| [caveman](https://github.com/JuliusBrussee/caveman) | Output compression for agents. |\n| [caveman-code](https://github.com/JuliusBrussee/caveman-code) | Terminal coding agent built around token efficiency. |\n| [cavemem](https://github.com/JuliusBrussee/cavemem) | Cross-agent memory. |\n| [cavekit](https://github.com/JuliusBrussee/cavekit) | Spec-driven build loop. |\n| [cavegemma](https://github.com/JuliusBrussee/finetune-caveman) | Fine-tuned model experiments for terse agent output. |\n\n## Skill Reference\n\nRun:\n\n```bash\nnode scripts/list-skills.mjs\n```\n\nVerify frontmatter:\n\n```bash\nnode scripts/verify-skills.mjs\n```\n\n## Contributing\n\nSkills should be:\n\n- Small enough to read quickly.\n- Triggered by clear user intent.\n- Progressive: load only the extra files needed.\n- Specific about workflow and validation.\n- Free of repo-specific assumptions unless the skill says so.\n\n## License\n\nMIT.\n", "bytes": 9582, "sha": "4727ea874e7e86f75eb633097b3d35c874d6f0c5a5fed01515b8c19a6d6b2bdd", "repo_slug": "juliusbrussee/skills", "fonte": "repo", "truncated": false, "api": "https://agentalog.com/api/listings/skl_juliusbrussee_skills_caveman_ee8072f0/readme" }