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.
git clone --depth 1 https://github.com/Kastalien-Research/thoughtboxWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/kastalien-research/thoughtbox/multi-agent-tdd)<a href="https://agentmods.dev/commands/kastalien-research/thoughtbox/multi-agent-tdd"><img src="https://agentmods.dev/badge/commands/kastalien-research/thoughtbox/multi-agent-tdd/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/kastalien-research/thoughtbox/multi-agent-tdd"><img src="https://agentmods.dev/badge/commands/kastalien-research/thoughtbox/multi-agent-tdd.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 5.1 | $0.00028 | $0.01097 |
| Opus 5.5 | $0.00011 | $0.00439 |
| Sonnet 5.5 | $0.00006 | $0.00219 |
| Haiku 4.5 | $0.00003 | $0.00110 |
Grade A, and why
multi-agent-tdd 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 1mo 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 — 110 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/multi-agent-tdd
Strict TDD workflow for hardening Thoughtbox as "Git for reasoning" — content-addressable thought hashing (SHA-256 Merkle chain), agent attribution on all thought operations, formal logic notation for conflict detection, and structural branch diffs with human-legible views.
Variables
RESUME: $ARGUMENTS (default: auto-detected from .multi-agent-tdd/state.json)
Module Order (dependency-driven)
| # | Module | Test Count | Type | Depends On |
|---|---|---|---|---|
| 1 | thought-attribution |
10 | Hybrid | — |
| 2 | content-hash |
10 | New (TDD) | M1 |
| 3 | cipher-logic-extension |
5 | New (TDD) | — |
| 4 | claim-parser |
9 | New (TDD) | M3 |
| 5 | conflict-detection |
8 | New (TDD) | M4, M1 |
| 6 | thought-diff |
10 | New (TDD) | M1, M4, M5 |
| 7 | identity-resilience |
6 | Validate | — |
| 8 | multi-client-demo |
9 | Integration | M1–M7 |
| Total | 67 |
Execution Order
M3 + M7 (parallel, no deps) → M1 → M2 + M4 (parallel) → M5 → M6 → M8
Anti-Drift Rules
- Red gate for new code: Every new module (M2-M6, M8) must have tests verified failing before implementation
- Hypothesis gate for existing code: M1, M7 tests should pass immediately against existing code
- Backward compatibility: Adding optional fields to ThoughtData must not break existing tests
- Regression after each module: All prior tests still pass
- Commit at every gate: Atomic commits for bisection
- State before code: Update
.multi-agent-tdd/state.jsonbefore and after each step
State File
Path: .multi-agent-tdd/state.json (gitignored)
Workflow
On Start
- Read
.multi-agent-tdd/state.json— if exists andstatus != completed, resume from last step - If no state file, initialize fresh state and begin at Phase 1
Per Module (New Code — M2-M6, M8)
pending→ Write tests →tests_written- Run tests → Verify red →
tests_red_verified - Implement →
implemented - Run tests → Verify green →
tests_green - Run regression →
complete - Commit
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.
- 1mo ago First seen · 110 lines · 28 tokens per session scan A b2d6c9f0020b
multi-agent-tdd is a command published in the GitHub repository Kastalien-Research/thoughtbox (66 stars, last pushed 3mo ago), licensed MIT. It adds 28 tokens to every session and 1,097 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-09-03.
Other commands, from other repositories
start
Start a new pair programming session with AI assistance.
tdd
Test-driven development with TodoWrite planning and comprehensive testing.
bugfix
Bug fix workflow: root cause analysis → user review → regression test + fix via TDD.
usage-add
PitWay: Accumulate measured planning or qa token usage onto a milestone.
complete
Complete a partially implemented feature by filling gaps and ensuring production readiness.
speckit.tdd.plan
Derive the feature's test list by delegating to zfa tdd plan (deterministic engine); falls back to LLM-guided derivation only when zfa is unavailable.