challenge
Skill24 starsChanged 19 days agoMITView on GitHub
Take on a hands-on RAG challenge
34 other projects use a version of this.
Adopted by
Most recently taken up by
- pantheon-org/tekhnetheir own version, first seen Nov 15, 2026
- yarikleto/claude-teacher-plugintheir own version, first seen Nov 14, 2026
- grainulation/grainulatortheir own version, first seen Nov 12, 2026
- pche-broken-artist/forge-of-thoughttheir own version, first seen Nov 10, 2026
- SylphxAI/skillstheir own version, first seen Nov 8, 2026
- seankoji-com/.agentstheir own version, first seen Nov 8, 2026
Used with
Projects with it also use prove-it, ingest and 8 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
Neko-Catpital-LabsSkillprove-it
Shared evidence rule for investigating, diagnosing, debugging, proving, or explaining what happened or why for any problem or situation: do not assert something is fixed, working, passing, merged, or running unless you personally observed it this turn and can show what you saw; do not assert it was caused by X until a repro and controlled isolation prove that cause.
Also in 3 projects that use this
JaimeGBASkillingest
Digest an external source (article, paper, transcript, URL, book chapter) and integrate its content into the knowledge base
Also in 4 projects that use this
GoogleSkillcritique
Expertise in auditing and fixing repository scripts and GitHub Actions workflows to ensure technical robustness and security.
Also in 3 projects that use this
genomewalkerSkilllearn
Learn from web sources, generate SKILL.md, and store key patterns in soul memory. Trigger keywords: learn, skill-gen, web-learn, documentation, tutorial.
Also in 4 projects that use this
Hugging FaceSkillimplement
Make tests pass. Invoke after /write-tests produces failing tests.
Also in 5 projects that use this
BlockSkillplan
Draft a TDD, PRD, or lightweight plan in the conversation without committing it.
Also in 4 projects that use this
Other versions34 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 34
- pedrohcgs/claude-code-my-workflow★ 1.7ksame name
- me2resh/apexyard★ 542same name
- nudgebee/nudgebee★ 525same name
- The01Geek/prflow★ 100same name
- grainulation/grainulator★ 86same name
- BULDEE/ai-craftsman-superpowers★ 35same name
- Hassaan146/claude-skills★ 18same name
- yarikleto/claude-teacher-plugin★ 15same name
- pche-broken-artist/forge-of-thought★ 14same name
- lenar-amirov/product-pipeline-public★ 13same name
24 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository23 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- architectureSkill
- audit-contentSkill
- benchmarkSkill
- break-itSkill
- buildSkill
- code-reviewSkill
- compareSkill
- debug-ragSkill
- evaluateSkill
- explainSkill
- fixSkill
- glossarySkill
- journalSkill
- lessonSkill
- paper-reviewSkill
- quizSkill
- recapSkill
- roadmapSkill
- sandboxSkill
- startSkill
- triageSkill
The file
We split this skill into the rules it gives an agent
7 rules. 0 are common, 6 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 30 edited versions we read keep these.
6
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 16
If the user specifies a difficulty or topic (e.g., /challenge advanced or /challenge multilingual), filter accordingly.
- Distinctive: line 25
Conversational RAG: Add conversation history to a RAG pipeline so follow-up questions work correctly (e.g., "What about the second one?").
- Distinctive: line 48
Test queries and expected behavior
- Distinctive: line 63
Do not build the solution for them.
- Distinctive: line 67
When the learner says they are done, evaluate their solution:
- Distinctive: line 71
Run the test queries and check the results
- Distinctive: line 104
Keep time expectations reasonable: beginner (1-2 hours), intermediate (2-4 hours), advanced (4-8 hours)
“Common” is measured: kept by at least 6 in 10 of the 30 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Challenge: Real-World RAG Challenges
Scope: This skill presents open-ended, multi-component challenges that require design decisions. For step-by-step guided construction of individual components, use
/build.
Present the learner with a realistic RAG challenge that tests their skills across multiple components. Challenges are more open-ended than /build exercises — they require design decisions, not just implementation.
Language awareness: See
.claude/LANGUAGE_AWARENESS.md.
Step 1: Select a Challenge
If the user specifies a difficulty or topic (e.g., /challenge advanced or /challenge multilingual), filter accordingly. Otherwise, recommend a challenge based on their progress.
Beginner Challenges
- FAQ Bot: Build a RAG system that answers questions from a FAQ document (provided). Must handle exact and paraphrased questions.
- Recipe Finder: Index a collection of recipes and build a retriever that finds recipes by ingredients, cuisine, or dietary restrictions.
- Book Q&A: Load a public-domain book, chunk it, and build a system that can answer questions about characters, plot, and themes.
Intermediate Challenges
- Multi-Document Synthesis: Build a RAG system that can answer questions requiring information from multiple documents. Test with a set of related articles.
- Conversational RAG: Add conversation history to a RAG pipeline so follow-up questions work correctly (e.g., "What about the second one?").
- Metadata-Filtered Search: Build a RAG system for product documentation where users can filter by product version, category, and date.
- Citation Generator: Build a RAG system that not only answers questions but provides exact source citations with page/paragraph references.
Advanced Challenges
- Multilingual RAG: Build a system that handles documents in multiple languages and queries in any of those languages.
- Legal Document Analyst: Build a RAG system for legal contracts that can answer questions about specific clauses, compare terms across contracts, and flag potential issues.
- Code Documentation Assistant: Build a RAG system over a codebase that can answer questions about architecture, find relevant functions, and explain code patterns.
- Real-Time RAG: Build a system that handles a continuously updating corpus (e.g., news articles) with minimal re-indexing latency.
Step 2: Present the Challenge
For the selected challenge, provide:
Challenge Brief
- Objective: What the learner needs to build (1-2 sentences)
- Requirements: Specific functional requirements (bulleted list)
- Constraints: Any limitations (e.g., must use open-source models, latency under 2 seconds)
- Evaluation criteria: How the solution will be judged
Provided Resources
- Sample data or instructions for obtaining data
- Suggested directory structure
- Test queries and expected behavior
Show a deliverable example in the learner's chosen language with proper types, documentation, and structure.
Hints (available on request)
Prepare a set of progressive hints — do not reveal them unless the learner asks. Start with high-level guidance and get more specific.
Step 3: Support During the Challenge
While the learner works, be available to:
- Answer clarifying questions about the requirements
- Provide hints if they get stuck (but encourage them to try first)
- Help debug specific technical issues
- Discuss design trade-offs when they face a decision
Do not build the solution for them. Guide them to find the answer.
Step 4: Evaluate the Solution
When the learner says they are done, evaluate their solution:
Functionality Check
- Does it meet all the stated requirements?
- Run the test queries and check the results
Design Review
- Is the architecture well-suited to the challenge?
- Were chunking, embedding, and retrieval choices appropriate?
- Is the prompt template effective?
Quality Assessment
Rate the solution on:
- Correctness: Does it produce accurate answers? (1-5)
- Robustness: Does it handle edge cases? (1-5)
- Code quality: Is the code clean and well-organized? (1-5)
- Design decisions: Were trade-offs made thoughtfully? (1-5)
Feedback
Provide specific feedback on what was done well and what could be improved. Suggest one stretch goal for extra credit.
Step 5: Track and Celebrate
Save the challenge completion to progress/challenges.md with the date, challenge name, difficulty, and score. If this is their first challenge at a given difficulty level, acknowledge the milestone.
Suggest 2-3 relevant next steps using slash commands:
/code-review— get expert feedback on the solution you just built/evaluate— measure the quality of your challenge solution with RAG metrics/roadmap— check your overall progress and see what to tackle next
Guidelines
- Challenges should feel achievable but require real thought and effort
- The learner should make design decisions, not just follow steps
- Accept multiple valid approaches — there is rarely one right answer
- Be encouraging throughout — challenges are meant to stretch, not discourage
- Keep time expectations reasonable: beginner (1-2 hours), intermediate (2-4 hours), advanced (4-8 hours)
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

