Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/uberumbra-caveman
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions: not checked yet.
What's inside
- Skill name
caveman- Description
- None
- License
- Apache-2.0
- Instructions
- 41 lines in SKILL.md · view on GitHub
Agent Skills spec issues
- invalid YAML frontmatter: Nested mappings are not allowed in compact mappings at line 2, column 14:
- name: missing
- description: missing
Files (1)
- SKILL.md
SKILL.md (first lines)
# Caveman Mode Terse communication mode for the atelier-pipeline. Reduces token usage while preserving technical accuracy. ## Core Rules 1. **Answer first.** No preamble, no greeting, no "Great question!". Start with the answer. 2. **Fragments OK.** "Config updated." beats "I have updated the configuration." 3. **20-word max sentences.** Long sentences split. 4. **Drop articles when meaning stays clear.** "Config updated" not "The configuration was updated." 5. **Active voice.** "Colby will fix" not "The fix will be handled by Colby." 6. **No hedging.** "Routing to Colby." not "I think we should probably route this to Colby." 7. **No closers.** No "Let me know if you need anything else." No "Happy to help!" 8. **Never omit negations.** "Not done" stays "not done". Never compress away a negative. 9. **Technical terms exact.** Never abbreviate or paraphrase technical identifiers, file paths, or error messages. ## Pipeline-Specific Applications - **Eva routing announcements:** one line, no ceremony. "Colby — config fix. Alternative: Poirot review first, skipped — no existing code to review." - **Poirot findings:** format per caveman-review skill. - **Agent status updates:** one sentence per beat. "Colby done. Test gate running." - **Phase transitions:** state the transition, not the explanation. "Sarah — ADR phase." ## Persistence Active for the session until user says "stop caveman", "normal mode", or "/caveman off". Set `caveman_mode: true` in `pipeline-config.json` to persist across sessions (requires Eva session-boot integration — ships in Unit 4 of ADR-0062). ## Break Rules For - Security warnings (must be complete and unambiguous) - Irreversible actions (deletion, force-push, production deploys — full sentence required) - Step-order sequences where partial omission causes incorrect execution ## Pre-Send Check Before each response in caveman mode: 1. Remove any sentence that announces what you're about to say. 2. Verify negations are present where needed. 3. Verify code snippets are accurate — never compress code.
From UberUmbra/atelier-pipeline, Apache-2.0.
More from UberUmbra/atelier-pipeline
- 0.561score
lean-build
UberUmbra
Strict scope enforcement for build tasks.
SkillAFreshness: AClaude Code +9 - 0.561score
surgical-patch
UberUmbra
Fix bugs at the narrowest responsible layer.
SkillAFreshness: AClaude Code +9 - 0.561score
ultracave
UberUmbra
Maximum compression mode.
SkillAFreshness: AClaude Code +9 - 0.561score
verify-and-stop
UberUmbra
Translate acceptance conditions into smallest sufficient proof set.
SkillAFreshness: AClaude Code +9 - 0.561score
investigate-first
UberUmbra
Separate observed symptom from inferred cause before editing.
SkillAFreshness: AClaude Code +9 - 0.558score
caveman-commit
UberUmbra
Terse commit messages.
SkillAFreshness: AClaude Code +9 - 0.555score
pipeline-setup
UberUmbra
Use when users want to install or set up the atelier-pipeline multi-agent orchestration system in their project.
SkillAFreshness: AClaude Code +9 - 0.474score
brain-hydrate
UberUmbra
Use when users want to bootstrap the brain with existing project knowledge -- reading ADRs, feature specs, UX docs, commit history, and error patterns to seed…
SkillCFreshness: CClaude Code +9 - 0.474score
brain-uninstall
UberUmbra
Use when users want to remove or disconnect the mybrain plugin's persistent memory layer.
SkillCFreshness: CClaude Code +9 - 0.409score
pipeline-uninstall
UberUmbra
Use when users want to remove the atelier-pipeline multi-agent orchestration system from their project.
SkillDFreshness: DClaude Code +9 - 0.409score
pipeline-uninstall
UberUmbra
Use when users want to remove the atelier-pipeline multi-agent orchestration system from their project.
SkillDFreshness: DClaude Code +9 - 0.304score
caveman-review
UberUmbra
No description from the publisher yet.
SkillAFreshness: AClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 9, 2026 (20 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- Apache-2.0
- Repository
- github.com/UberUmbra/atelier-pipeline
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from UberUmbra
- 0.561score
verify-and-stop
UberUmbra
Translate acceptance conditions into smallest sufficient proof set.
SkillAFreshness: AClaude Code +9 - 0.561score
ultracave
UberUmbra
Maximum compression mode.
SkillAFreshness: AClaude Code +9 - 0.561score
investigate-first
UberUmbra
Separate observed symptom from inferred cause before editing.
SkillAFreshness: AClaude Code +9 - 0.561score
lean-build
UberUmbra
Strict scope enforcement for build tasks.
SkillAFreshness: AClaude Code +9 - 0.561score
surgical-patch
UberUmbra
Fix bugs at the narrowest responsible layer.
SkillAFreshness: AClaude Code +9 - 0.558score
caveman-commit
UberUmbra
Terse commit messages.
SkillAFreshness: AClaude Code +9
Cite this page
caveman skill. Agent Mart. https://agentmart.dev/skills/uberumbra-caveman (last updated 2026-10-24)