Stacker

Stack overlapping skills instead of picking one. When two or more installed skills or tools can do the same job (web scraping, search, code review, test generation, data extraction), run the best candidates in parallel on the same task, score their outputs, and merge them into one result that beats any single tool. Use when the user says "stack", "combine skills", "use every tool", "ensemble", "cross-check", "double-check with another tool", when a task is output-critical and multiple capable skills are installed, or when evals show no single tool wins on every axis. Also use when the user wan

Install

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add ogarciarevett/stacker --skill stacker
Manual install
git clone --depth 1 https://github.com/ogarciarevett/stacker.git /tmp/stacker
mkdir -p ~/.claude/skills
cp -r /tmp/stacker/skills/stacker ~/.claude/skills/stacker

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# stacker

Skills overlap. Run a minimal eval over your installed skills and you find the
uncomfortable truth: **no tool wins on every axis.** One scraper is faster, another
renders JavaScript, a third survives bot walls. One reviewer catches security holes,
another catches naming. The deltas are small and axis-specific.

Agents handle this badly in two symmetric ways:

- **Load everything.** Every overlapping skill goes into context "just in case."
  Context bloats, instructions conflict, the agent improvises.
- **Pick "the best."** A router chooses one winner and throws away every axis
  the losers were better at.

Stacker is the third option: **run the top candidates on the same frozen task,
score their outputs, merge them.** Ensembling, applied to agent skills. The unit
you combine is the *output*, never the *context*.

## When to stack (and when not to)

Stacking costs tokens, time, and sometimes money. It is the paid upgrade, not the
default. Stack only when ALL of these hold:

1. **Output-critical** — correctness or completeness matters more than latency and cost.
2. **Real overlap** — 2+ installed skills can each do the task end to end.
3. **Complementary failures** — evals or experience show they fail on *different* inputs.

| Signal | Action |
|---|---|
| Routine task, one obvious tool | Route (single tool, escalate on failure) |
| One tool strictly dominates every axis | Route |
| Output feeds a decision, a publish, or money | **Stack** |
Read the full skill on GitHub

Details

Skill name
stacker
Path
skills/stacker/SKILL.md
License
MIT
Repo updated
24 October 2026

FAQ

What does the Stacker skill do?

Stack overlapping skills instead of picking one. When two or more installed skills or tools can do the same job (web scraping, search, code review, test generation, data extraction), run the best candidates in parallel on the same task, score their outputs, and merge them into one result that beats any single tool. Use when the user says "stack", "combine skills", "use every tool", "ensemble", "cross-check", "double-check with another tool", when a task is output-critical and multiple capable skills are installed, or when evals show no single tool wins on every axis. Also use when the user wan

How do I install the Stacker skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add ogarciarevett/stacker --skill stacker. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Stacker popular?

It lives in ogarciarevett/stacker, which has 9 GitHub stars and 0 forks.

Related skills

Xbrain

9

corespeed-io/xbrain

Organize authorized saved references into evidence-based design categories, visual features and tentative styles, then retrieve original images and sources for an Agent's design task. Use when someone has saved UI, graphic or interior references but cannot describe the style they want.

Productivity & Notes

Freelo

9

freeloio/claude-freelo-skill

Use when interacting with the Freelo.io project management API — managing projects, tasklists, tasks, subtasks, comments, time tracking, work reports, labels, custom fields, notes, files, templates, pinned items, workers, notifications, events, invoices, or searching in Freelo. Trigger this skill whenever the user mentions Freelo, wants to manage tasks or projects in Freelo, track time, create work reports, check invoices, or do anything related to Freelo.io — even if they do not explicitly say "API".

Productivity & Notes

Compliance Validator

10

gener8v/gener8v.claude-skills.rfp-response

Validates every reviewed section against the RFP's explicit requirements. Produces a pass/fail compliance report with remediation notes. Runs after the empathy reviewer because tone fixes can inadvertently remove compliance-critical language. Trigger when the user says "check compliance," "validate the sections," "are we compliant," "run the compliance check," or after the empathy reviewer has produced sections/*-reviewed.md files.

Productivity & Notes

Adhd Operating System

9

brkhzn/adhd-operating-system

Pervasive filter adapting all Claude interactions to inattentive ADHD cognition. Affects BOTH response formatting/structure AND content recommendations. Trigger aggressively for: productivity, organization, project management, task management, learning, studying, note-taking, communication, email, meetings, habits, routines, daily planning, time management, tool selection, workspace setup, career decisions, writing, creative work, prioritization, or anything involving executive function, working memory, attention, motivation, or emotional regulation. Also trigger for mentions of ADHD, focus, d

Productivity & Notes

Dataroom Prep

10

Founders-Edge/toolkit

Help a founder build and organize a fundraising dataroom before they start raising. Use this skill whenever someone asks what goes in a dataroom, to build or set up a dataroom, to organize their dataroom, what documents investors will ask for, or how to prepare for diligence. Triggers on: "prep my dataroom", "build a dataroom", "what goes in a dataroom", "organize my dataroom", "what docs do investors need", "get ready for diligence", "set up our data room". This is the build-side partner to the diligence-review skill: diligence-review finds the gaps, dataroom-prep assembles the folder and tel

Productivity & Notes

Memanto Cookbooks

9

moorcheh-ai/memanto-agent-skills

End-to-end blueprints for building memory-powered applications on MEMANTO — persistent agent memory, session continuity, memory export and audit, automated daily digests, conflict resolution, and memory-grounded RAG. Use when designing or scaffolding a project that needs agent memory, not for one-off remember/recall calls.

Productivity & Notes

Browse the most popular Claude skills, all productivity & notes skills, or write your own with the skills generator.