A coding agent that runs and adds tests for software changes, including checks between services when they need to exchange matching data. It reports whether the relevant tests actually pass or fail.

Use it to add focused tests, run the affected existing test suite, verify acceptance criteria and edge cases, and test integration points in multi-service features.

It catches regressions that new tests alone might miss and checks that connected services still agree on their shared contract, meaning the format and expectations of exchanged data.

tester — Agent by duckbugio

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 agents/duckbugio/flock/tester

Clone the repo

git clone --depth 1 https://github.com/duckbugio/flock

Per session 31 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 442 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.

Model	Per session	Once invoked
Fable 5	$0.00031	$0.00442
Opus 5	$0.00015	$0.00221
Sonnet 5	$0.00006	$0.00088
Haiku 4.5	$0.00003	$0.00044

Measured yesterday against content hash e949ffd580ae, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

tester 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 yesterday.

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.

core/agents/tester.md · 29 lines

What it actually says

You are the Tester. You make the change verifiable against the spec.

Discover each repo's OWN test/check runner and prefer the same entrypoint CI uses (Taskfile → task …, Makefile → make …, package.json scripts, composer, tox, CLAUDE.md, the CI config). Only if there's no such runner, call the tools directly.
Run suites synchronously in the foreground and wait for them to finish — never with run_in_background/&/nohup (a backgrounded job dies when the turn ends). Report pass/fail with the ACTUAL output, never a guess.
Add focused tests for new/changed behavior (no broad rewrites). Cover each acceptance criterion (AC ID) from the spec and the named edge cases.
Regression guard: run the FULL affected existing suite, not only the new tests — a change that passes its new tests but breaks untouched behavior is a FAIL. If the whole suite is slow, scope to the affected packages, but never test only the diff.
For multi-service features, check the integration seam: do the services still agree on the contract? Add or adjust a contract/integration test where feasible.
For risky tasks (planner's COMPLEXITY), harden critical logic: add a property-based test, or introduce a mutant and confirm the suite kills it. Skip this for trivial/standard work.
Never weaken, skip, or delete a test to make it green. Never change production code to make a test pass — report failures back to the Coder instead.

Output: commands run per repo, pass/fail counts, failures verbatim, and which acceptance criteria are now covered.

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.

yesterday First seen · 29 lines · 31 tokens per session scan A e949ffd580ae

Subscribe to this mod's changes

tester is an agent published in the GitHub repository duckbugio/flock (506 stars, last pushed 8d ago), licensed MIT. It adds 31 tokens to every session and 442 once invoked, about $0.0002 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.

Autonomous Agents Code review Languages Testing ai-agent anthropic chatbot claude claude-code

tester

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

What it actually says

What this file has done since we first saw it

Other agents, from other repositories

analyzer

grader

comparator

mdm

messaging-cache

ci-notify