semble
Skill5 starsChanged 3 months agoMITView on GitHub
Use when exploring an unfamiliar codebase, locating where functionality is implemented, or answering "where is X / how does Y work" questions. Run `semble search` instead of grep+read for semantic or exploratory queries — it returns the relevant code chunks directly and uses ~98% fewer tokens. Triggers: "where is", "how does", "find the code that", code exploration, feature lookup.
In this repository16 other files it gives its agents.
- agentsview-cliSkill
- clarifySkill
- deslop-biSkill
- eli5Skill
- engram-memSkill
- falsifySkill
- ghSkill
- git-commitSkill
- herdrSkill
- hunk-reviewSkill
- mcp2cliSkill
- opencliSkill
- show-meSkill
- tavilySkill
- tmuxSkill
- worktrunkSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 1 is distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 17
Prefer grep only for exhaustive literal matches or confirming an exact string.
- Very specific: line 41
If semble is not on $PATH, semble isn't installed in this environment.
- Very specific: line 53
Start with semble search to find relevant chunks.
- Very specific: line 54
Use --content docs / config / all when the answer isn't in code.
- Distinctive: line 55
Read full files only when the returned chunk lacks enough context.
- Very specific: line 56
Use semble find-related <file> <line> <path> on a promising result to discover related implementations.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Semble — Code Search via CLI
Fast semantic code search. Returns the exact chunks you need, ~98% fewer tokens than grep+read.
When to use
- Exploring an unfamiliar codebase / answering "where is X" / "how does Y"
- Locating the implementation of a feature, concept, or flow
- Finding code semantically similar to a known location
Prefer grep only for exhaustive literal matches or confirming an exact
string. Use Read after semble only when a returned chunk doesn't give
enough context.
Core commands
# Search by intent or symbol (path defaults to current dir; git URLs ok)
semble search "authentication flow" ./my-project
semble search "save_pretrained" ./my-project
semble search "save model to disk" ./my-project --top-k 10
# Search docs / config / everything (default is code only)
semble search "deployment guide" ./my-project --content docs
semble search "database host port" ./my-project --content config
semble search "authentication" ./my-project --content all
# Find code similar to a known location (file path + line number)
semble find-related src/auth.py 42 ./my-projectIndexes are built and cached on first run, and invalidated automatically when files change.
If semble is not on $PATH, semble isn't installed in this environment.
Tell the user and suggest installing it with:
uv tool install semble(Requires uv.) Until it's installed, fall
back to grep + Read.
Workflow
- Start with
semble searchto find relevant chunks. - Use
--content docs/config/allwhen the answer isn't in code. - Read full files only when the returned chunk lacks enough context.
- Use
semble find-related <file> <line> <path>on a promising result to discover related implementations. - Fall back to
greponly for exhaustive literal matches or exact-string confirmation.
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.

