read-the-docs
Skill0 starsChanged 4 days agoMITView on GitHub
Look things up in the local docs, then the skills, then the example, then the code, before relying on memory. Use when you are unsure how an action, approval, lint rule, the doctor, the CLI, storage or a model provider works, or when about to guess an API.
In this repository29 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- accessibilitySkill
- actionsSkill
- adding-a-featureSkill
- agent-workflowSkill
- app-contextSkill
- approval-and-safetySkill
- backend-proxySkill
- background-workSkill
- build-an-appSkill
- capture-learningsSkill
- clean-architectureSkill
- customizing-the-assistantSkill
- delegate-to-agentSkill
- edit-boundariesSkill
- i18n-and-rtlSkill
- model-providersSkill
- performanceSkill
- permissions-and-platformSkill
- reliable-mutationsSkill
- routingSkill
- secretsSkill
- state-managementSkill
- storing-dataSkill
- testingSkill
- ui-designSkill
- upgrade-the-frameworkSkill
- verifySkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 8
Don't load this for saving what you learned.
- Very specific: line 28
Read the example: If the project has a .agents/example/ folder, it holds the item example as reference, mirroring lib/ and test/:
- Very specific: line 52
No automated check exists for this behaviour.
≈ “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
Read the docs
Don't load this for saving what you learned; see capture-learnings.
Rule
Before you answer from memory, look it up in this order: the local docs, then the skill for the topic, then the example folder when the project has one, then the code. The docs come with the installed version of the framework, so no web search is needed for the framework's own behaviour.
Why
This repo's conventions differ from the generic Flutter ones, and the framework changes between versions. A remembered API that was renamed, or a rule that no longer exists, produces code that fails the gate. The docs are short and local.
How
- Search the docs.
dart run flutter_agentify_core:docs --list dart run flutter_agentify_core:docs --query "needs approval"--listprints the page names inpackages/flutter_agentify_core/doc/:actions.md,agent-loop.md,app-context.md,approval.md,cli-and-doctor.md,flutter-widgets.md,lints.md,model-providers.md,storage-and-memory.md.--queryprints<file>:<line>: <text>for each line holding every word, ignoring case. It exits with 1 when nothing matches, so try fewer or different words. - Open the page. For actions, read
packages/flutter_agentify_core/doc/actions.md(theAgentActionshape, flags, failure codes such asapproval_required). - Load the skill in
.agents/skills/for the topic.AGENTS.mdgroups them by area. - Read the example. If the project has a .agents/example/ folder, it holds the item example as reference, mirroring
lib/andtest/:.agents/example/lib/domain/actions/items_actions.dart<!-- @empty: Open the file for the same layer as your task; it is a reference, not app code. --> - Read the code. Copy a real example:
lib/domain/actions/items_actions.dartfor actions,lib/ai/screen_contexts.dartfor screen context. The code beats the docs if they disagree; mention the difference in your summary.
Look up when:
- you are about to use a framework class or method you have not seen in this repo;
- a lint rule or doctor check fails and you do not know the rule;
- you are writing an action, an approval, a storage change or a provider;
- a command or flag is needed (
tool/action.dart,doctor,docs); - the task mentions something the repo may already have.
Anti-patterns
- Writing an
AgentActionfrom memory of another project. - Searching the web for how the framework behaves, when the local pages answer it.
- Giving up after one empty
--query; the search needs every word on the same line. - Quoting a doc without checking the code that uses it.
- Reading all nine pages when one query finds the line.
Verification
No automated check exists for this behaviour. The skills doctor check only confirms a skill's frontmatter, size and repo paths. You can show you looked up by naming the page or file you used.
Related skills
actions— how this repo writes an action.agent-workflow— when a lookup does not settle it and you must ask.capture-learnings— saving a surprising result of a lookup.
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.

