adk-eval

A testing assistant for Botpress ADK bots. An eval is a repeatable check that compares a bot's behavior with expected results.

In plain words
What is it for?
Use it to find existing evals, write tests for bot primitives, run them, investigate failures, and inspect execution traces.
Why use it?
It helps reveal untested bot features and distinguish problems in the bot from problems in the test's expectations.

Command

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add commands/botpress/skills/adk-eval
Clone the repo
git clone --depth 1 https://github.com/botpress/skills
Per session 18 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 272 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00018 $0.00272
Opus 5 $0.00009 $0.00136
Sonnet 5 $0.00004 $0.00054
Haiku 4.5 $0.00002 $0.00027

Measured 2d ago against content hash 381cb0545323, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

adk-eval scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

commands/adk-eval.md · 18 lines

What it actually says

Load the adk-evals and adk skills, then help with evals immediately.

If $ARGUMENTS names a primitive or feature (e.g., "search", "checkout", "tickets"), check if evals already exist (Glob({ pattern: "evals/**" })). If found, ask: run them, debug them, or write a new one? If not found, search src/ for related primitives, read the source code, and generate an eval tailored to that code with realistic assertions.

If $ARGUMENTS asks to run evals, run adk evals with appropriate flags.

If $ARGUMENTS mentions a failure ("failing", "broken", "wrong"), inspect the failing eval's traces and identify whether the issue is the eval's assertions or the bot's behavior.

If $ARGUMENTS is empty, check existing eval coverage (Glob({ pattern: "evals/**" })) and the user's primitives in src/, then suggest writing evals for untested primitives.

Follow the eval format, assertion types, and per-primitive testing patterns from the skill documentation.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 18 lines · 18 tokens per session scan A 381cb0545323

Subscribe to this mod's changes

adk-eval is a command published in the GitHub repository botpress/skills (11 stars, last pushed 1mo ago), licensed MIT. It adds 18 tokens to every session and 272 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.