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.
npx agentmods add skills/kelp/agent-plugins/tdd-orchestratenpx skills add kelp/agent-plugins --skill tdd-orchestrategit clone --depth 1 https://github.com/kelp/agent-pluginsWhat 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 | $0.00143 | $0.04158 |
| Opus 5 | $0.00072 | $0.02079 |
| Sonnet 5 | $0.00029 | $0.00832 |
| Haiku 4.5 | $0.00014 | $0.00416 |
Grade A, and why
tdd-orchestrate 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.
How it starts
The opening of the file, as written. The whole thing — 498 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/tdd-orchestrate
Strict red-green-refactor for every code change.
Input
Task: $0
If no task was provided, ask the user what to implement or fix. One sentence is enough.
Model
Agents inherit your session model by default. Cheaper models tend to fail the reviewer gates and loop, which costs more wall-clock and tokens than a capable model one-shotting the stage — so prefer a capable model for this pipeline.
To pin a model regardless of session, pass --model <name> in the task (e.g. /tdd-orchestrate --model opus parser). Strip the flag from the task text before
briefing agents, then pass model: <name> to every
Agent dispatch in this skill (inline and full pipeline
alike). With no flag, omit model: so each agent
inherits the session model.
Decide: Pipeline or Inline
A behavior is one discrete, independently testable unit of functionality the module must exhibit (e.g. "the parser rejects an empty string," "the cache evicts the oldest entry"). Every later reference to "behavior" in this skill and in the agent role files means this definition — one item in the orchestrator's enumerated behavior list for the task.
Use the full pipeline (see "Pipeline" section) when:
- Building a new module from scratch
- The task has 3+ distinct behaviors to implement
- The user explicitly asks for the pipeline
Use inline red-green-refactor when:
- Fixing a bug
- Adding a single feature or flag
- Modifying existing code
- The change touches 1-2 files
Most tasks are inline. Default to inline unless the scope clearly warrants the full pipeline. Precedence: if any full-pipeline criterion applies, use the full pipeline regardless of the file-count heuristic — a new module that happens to fit in one file is still a full-pipeline task.
Inline Red-Green-Refactor
Precondition: read the project's CLAUDE.md. If it
has no ## TDD Pipeline Configuration section, stop and
tell the user to run /tdd-pipeline:tdd-init first.
Inline uses two agents and two commits. It skips the reviewer stages and the stub/RED-gate dance, because inline targets existing code where the test fails against the bug directly (no stub to typecheck against).
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.
- 2d ago First seen · 498 lines · 143 tokens per session scan A 5ab6352f9c85
tdd-orchestrate is a skill published in the GitHub repository kelp/agent-plugins (2 stars, last pushed 9d ago), licensed MIT. It adds 143 tokens to every session and 4,158 once invoked, about $0.0007 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 2026-08-31.
Other skills, from other repositories
test
Generate tests and coverage plans. Triggers: "test", "generate tests and coverage plans.", "test skill".
auto-loop
TDD-based autonomous development loop with checkpoint recovery and observability changelog.
test-first
Drive one feature through a strict TDD Red-Green-Refactor cycle with checklists for each phase. Use when implementing new functionality test-first, or when the user runs /test-first.
feature
End-to-end feature workflow around the agent — spec by interview, a plan of 2–5-minute tasks, test-first execution with checkpoint commits, then an evidence-gated finish. Effort-scaled: a one-sentence diff skips straight to implementation; a real feature gets spec.md → plan.md → task-by-task TDD, with all state in…
testing-go
Writes tests following TDD (using go test, testify, and rapid) best practices. Use when writing unit tests, integration tests, or table-driven tests in Go.
agent-integration
Run all three agent integration phases sequentially: research, write-tests, and implement using E2E-first TDD (unit tests written last). For individual phases, use /agent-integration:research, /agent-integration:write-tests, or /agent-integration:implement. Use when the user says "integrate agent", "add agent…