agentleFS
samwelmungasamwelmunga/jenga-npm/.agents/skills/jenga-config/SKILL.md

j.jenga-config

Skill0 starsChanged 23 days agoApache-2.0View on GitHub

List, read and change Jenga's project settings (project/configs/*.json) through the deterministic `jenga config` command. A thin wrapper - it collects the file, key and value in chat and runs one script; the edit itself is validated and written by code, with no model reasoning.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    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
  1. Very specific:

    Run it from the project root (the command finds project/configs/ from the working directory).

    line 38
  2. Distinctive:

    If the user did not already say, ask:

    line 44
  3. Distinctive:

    Read one setting (get)

    line 48

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

jenga-config - Read or Change Jenga Settings

Purpose

/jenga-config is a thin chat front end for jenga config, the command that lists and edits the settings in project/configs/*.json. The skill only collects the inputs and reports the result. Validation, type and range checks, the version-counter bump and the atomic write all happen inside the CLI, so an edit costs no model reasoning beyond gathering three inputs.

It is a main skill like /jenga and /jenga-permission-level: its directory keeps the bare name skills/jenga-config/ (one of the permanent exceptions in the Canonical Naming Contract), and it is invoked as /jenga-config or j.jenga-config.

Hard rules

  1. One entry point. Every action runs run-config.sh and nothing else. Never read, parse, edit or write project/configs/*.json yourself, and never call jenga or node bin/jenga.js directly. The script resolves the CLI (a jenga on PATH, otherwise a pinned copy installed into a per-user cache), so the CLI does not need a global install.
  2. Writes are committed project values. A set changes the file in project/configs/, which is version-controlled and shared with the whole project. There is no local-override layer and no --shared flag. Do not offer either.
  3. Report verbatim. Show the script's stdout and its exit code exactly as returned. Do not paraphrase, summarise, or "fix up" the output.

The script

SCRIPT="$([ -f skills/jenga-config/scripts/run-config.sh ] && echo skills/jenga-config/scripts/run-config.sh || echo node_modules/@jenga-ai/agent/skills/jenga-config/scripts/run-config.sh)"
bash "$SCRIPT" <subcommand> <args>

Run it from the project root (the command finds project/configs/ from the working directory).

Instructions

1. Choose the action

If the user did not already say, ask:

What would you like to do?
1. List the settings and edit interactively
2. Read one setting (get)
3. Change one setting (set)
4. Other (describe below)

2. Gather only what the action needs

An address is <file-id>.<key>, where <file-id> is a config file's name without .json (for example scope-thresholds.inline_max_files). If the user does not know the address, run the list action first and let them pick; do not open the config files to look it up.

ActionCollectRun
Listnothingbash "$SCRIPT"
Getthe addressbash "$SCRIPT" get <file-id>.<key>
Setthe address and the new valuebash "$SCRIPT" set <file-id>.<key> <value>

Quote the value if it contains spaces or shell characters. For a set, state the address and value once and run it; the CLI rejects an invalid value without writing anything, so no extra confirmation round is needed beyond the user having just supplied the value.

The list action starts the CLI's interactive picker. If the shell has no terminal and it ends without output, say so and tell the user to run jenga config in a terminal, or to use get / set with an address.

3. Report

Print the script's stdout and the exit code verbatim. Exit codes:

CodeMeaning
0Success (a set of the value already in the file reports (unchanged)).
2Usage error: missing or surplus arguments.
3Unknown config file or unknown key.
4Invalid value: wrong type, out of range, not an allowed value.
5Read-only key: nothing was written.
6Config file missing or malformed, or the configs directory could not be resolved.
7The jenga CLI could not be installed. The config command was not run. Show the script's stderr.

For a non-zero code, show the stderr the script printed and stop; do not retry with a different value on the user's behalf.

4. After a successful set

The file under project/configs/ has changed on disk. Tell the user it is an uncommitted project change and that j.commit records it. Do not commit it yourself.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.