agentleFS
GoogleKnown publishergoogle/adk-samples/core/python/ambient-quality-agent/AGENTS.md

adk-samples / ambient-quality-agent

AGENTS.md11k starsChanged 44 days agoApache-2.0View on GitHub

These rules apply to comments, docstrings, and user-visible messages (logs, printed output).

1 other project uses a version of this.

Trending

More talk lately.about the whole repository

Dec 2025: 1 mentionJan 2026: 1 mentionFeb 2026: 2 mentionsMar 2026: 0 mentionsApr 2026: 0 mentionsMay 2026: 1 mentionJun 2026: 1 mentionJul 2026: 2 mentionsAug 2026: 0 mentionsSep 2026: 4 mentionsOct 2026: 4 mentionsNov 2026: 0 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the google/adk-samples repository that our weekly crawl found. There is no adoption data for this file yet.

Most recently taken up by

Who's talking about google/adk-samples

The biggest 5 of 32 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 32

Talked about on GitHub, Bluesky, Hacker News and Stack Overflow (32 times).

On GitHub · 23

18 more on GitHub

On Bluesky · 4

The file

We split this AGENTS.md into the rules it gives an agent

7 rules. About 0 are common, 4 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    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
  1. Distinctive:

    Every function and method name starts with a verb that states what it does.

    line 5
  2. Very specific:

    Do not use generate_, process_, do_, make_, or handle_ (except UI/event handlers).

    line 31
  3. Distinctive:

    Write plain, professional English that a reader understands on first pass.

    line 41
  4. Distinctive:

    Use a numbered list to describe a sequence of steps, so each step maps to the code that performs it.

    line 46
  5. Very specific:

    Every function docstring documents its arguments and return value in Google style (Args:, Returns:, Raises: where relevant).

    line 47
  6. Distinctive:

    Every claim must be true of the code.

    line 48
  7. Very specific:

    When the pr-description subagent is available, write every PR description with it, so all PRs share one format.

    line 57

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md 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
AGENTS.md

ambient-quality-agent: agent instructions

Naming

  • Every function and method name starts with a verb that states what it does. This includes pure helpers.
  • Exempt: @property members (nouns), dunders, pytest test_* functions and fixtures, framework overrides (OpenTelemetry shutdown/force_flush, ADK callbacks such as before_agent).
  • Mirrored code shares names:
    • the BigQuery stores and readers under src/ambient_quality_agent/tools/ ↔ their in-memory counterparts in src/ambient_quality_agent/standalone/ (InMemory*), which implement the same contract;
    • TrajectoryReader/InsightReader ↔ tools/mock_aqa_server.py.

Verbs

VerbUse for
get_Fetch one item by key, returning X | None; or several items by keys, returning a dict keyed by id
list_Return a collection, optionally filtered or paginated
count_, sum_Aggregate queries
read_, write_, store_Raw object, file or request-body I/O
load_Read and parse into a typed object
fetch_Telemetry ingestion paging and remote HTTP only
save_Upsert an entity
record_, append_Append to an event log or a table
insert_Raw BigQuery insert
create_Create a stored resource, or a default dependency (_create_default_<thing>)
build_Assemble an in-memory object, including SQL fragments and URLs
render_Turn data into display text (markdown, report lines)
format_Reshape one value or record for output
parse_, decode_Text to structure; decode an encoded token
compute_, resolve_, extract_Derive a value; choose among alternatives; pull a field out of an argument
  • Do not use generate_, process_, do_, make_, or handle_ (except UI/event handlers).
  • Yes/no functions start with is_, has_, can_ or should_, or use a verb that reads as a question (matches_, covers, defines_).
  • Conversions: to_<x> (methods), from_<x> (alternate constructors), <src>_to_<dst> (free functions). Do not use as_<x>.
  • CLI command functions are named <verb>_<noun>_cmd.
  • A name must not shadow a builtin or another name in the same scope.

Comments and docstrings

These rules apply to comments, docstrings, and user-visible messages (logs, printed output).

  • Write plain, professional English that a reader understands on first pass. Use US spelling, unless a file consistently uses another convention.
  • Explain why, not what. Delete a comment that only restates what the code plainly does.
  • Be concise, but keep the reasoning a reader needs.
  • Describe only the current state. No historical or migration language ("previously", "no longer", "instead of", "now we", "used to").
  • No commented-out code.
  • Use a numbered list to describe a sequence of steps, so each step maps to the code that performs it.
  • Every function docstring documents its arguments and return value in Google style (Args:, Returns:, Raises: where relevant). pytest test functions and fixtures are exempt.
  • Every claim must be true of the code. Verify a stated reason before you keep or reword it; never invent one.

Module diagrams

  • Some module docstrings contain an ASCII wiring diagram (the literal block after a line ending in ::). When you change the wiring a diagram shows (a component, a call between components, or an implementation), update the diagram in the same change.
  • Diagrams show stable structure only: components, call direction and method names. No line numbers, parameters or config values.

Pull request descriptions

  • When the pr-description subagent is available, write every PR description with it, so all PRs share one format. Use its output as the PR body.

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.