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 commands/punt-labs/z-spec/b-create-devgit clone --depth 1 https://github.com/punt-labs/z-specWhat 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 | $0.00012 | $0.01909 |
| Opus 5 | $0.00006 | $0.00955 |
| Sonnet 5 | $0.00002 | $0.00382 |
| Haiku 4.5 | $0.00001 | $0.00191 |
Grade A, and why
b-create-dev 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.
Copies of this mod
1 near-identical copy found in the catalogue:
- b-create — 94% identical, 6 lines differ
How it starts
The opening of the file, as written. The whole thing — 315 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/z-spec-dev:b-create-dev - Create B Machine
You are creating a B Abstract Machine Notation (AMN) specification. B machines are executable formal specifications that can be type-checked, animated, model-checked, and refined toward implementation --- all using probcli.
This command has two modes:
- Description mode: Create a B machine from a natural language description
- Translation mode: Translate an existing Z specification (
.tex) to a B machine
Input
Hint: $ARGUMENTS
Determine the mode:
- If the argument is a path to a
.texfile: translation mode - Otherwise: description mode (treat as natural language description)
Process
0. Check Prerequisites
Verify probcli is installed:
PROBCLI="${PROBCLI:-$HOME/Applications/ProB/probcli}"
if ! which probcli >/dev/null 2>&1 && [ ! -x "$PROBCLI" ]; then
echo "PROBCLI_NOT_FOUND"
fi
If probcli not found: Stop and tell the user:
probcli is not installed. Run
/z-spec-dev:setup-dev probclifirst.
1. Create specs/ Directory
mkdir -p specs
Update .gitignore if needed:
for pattern in "specs/*.prob" "specs/*.prob2project"; do
grep -qxF "$pattern" .gitignore 2>/dev/null || echo "$pattern" >> .gitignore
done
2a. Description Mode
If working from a natural language description:
Analyze the Description
Extract:
- Entities: What things does the system track?
- Attributes: What properties do entities have?
- Constraints: What must always be true?
- Operations: What state transitions exist?
- Relationships: How do entities relate to each other?
Generate the Machine
Create a .mch file following this structure:
MACHINE MachineName
SETS
/* Deferred sets for entity identifiers */
ID;
/* Enumerated sets for status values */
STATUS = {active, inactive}
CONSTANTS
max_capacity
PROPERTIES
max_capacity : NAT &
max_capacity = 100
VARIABLES
/* State variables */
entries,
statuses
INVARIANT
/* Typing constraints */
entries : ID +-> NAT &
statuses : ID +-> STATUS &
/* Domain consistency */
dom(entries) = dom(statuses) &
/* Business invariants */
card(entries) <= max_capacity
INITIALISATION
entries := {} ||
statuses := {}
OPERATIONS
/* Operations with PRE/THEN substitutions */
add(id, val) =
PRE id : ID & val : NAT &
id /: dom(entries) &
card(entries) < max_capacity
THEN
entries := entries \/ {id |-> val} ||
statuses := statuses \/ {id |-> active}
END;
remove(id) =
PRE id : ID & id : dom(entries)
THEN
entries := {id} <<| entries ||
statuses := {id} <<| statuses
END;
nn <-- count =
nn := card(entries)
END
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 · 315 lines · 12 tokens per session scan A 2a9d8f4b37cb
b-create-dev is a command published in the GitHub repository punt-labs/z-spec (5 stars, last pushed 3d ago), licensed MIT. It adds 12 tokens to every session and 1,909 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 2026-08-31.
Other commands, from other repositories
cross-language-check
Replicate a quantitative analysis in a second language (R↔Python↔Stata↔Julia) and compare outputs for implementation errors. Use when an existing empirical result needs independent cross-language verification. Not for reviewing one implementation in place; use $code-suite.
lean-check
Formalize a self-authored lemma or theorem in Lean 4/mathlib and require a clean lake build without sorry. Use when the mathematical claim can be stated faithfully and machine-checked. For numerical falsification or symbolic algebra, use $numerical-check or $symbolic-check.
experiment-design
Design empirical studies through power analysis, pre-analysis planning, QSF parsing, and survey architecture. Use when specifying sampling, measurement, treatment, or analysis before data collection. Not for causal identification alone; use $causal-design.
feedback
Incorporate feedback into PR/FAQ and redraft affected sections.
causal-design
Design or audit the identification strategy for an observational study. Use when the task concerns estimands, causal assumptions, threats to identification, or defensible research design rather than model implementation.
computational-experiments
Scaffold, execute, analyse, and publish computational research experiments through a reproducible staged workflow. Use when a research question requires simulations or computational sweeps rather than a one-off script.