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 instructions/hpi-swa/trufflesqueak/agents-mdgit clone --depth 1 https://github.com/hpi-swa/trufflesqueakWrote 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/instructions/hpi-swa/trufflesqueak/agents-md)<a href="https://agentmods.dev/instructions/hpi-swa/trufflesqueak/agents-md"><img src="https://agentmods.dev/badge/instructions/hpi-swa/trufflesqueak/agents-md.svg" alt="Measured on agentmods" 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 | $0.01043 | $0.01043 |
| Opus 5 | $0.00522 | $0.00522 |
| Sonnet 5 | $0.00209 | $0.00209 |
| Haiku 4.5 | $0.00104 | $0.00104 |
Grade B, and why
trufflesqueak AGENTS.md scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
sudo apt-get update How it starts
The opening of the file, as written. The whole thing — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.
TruffleSqueak Agent Reference
Preflight Checks
Run these before spending time on build failures:
java --version
mx --version
python3 --version
cmake --version
python3 -c "import re; print(re.search(r'\"version\"\s*:\s*\"(\d+\.\d+\.\d+)', open('mx.trufflesqueak/suite.py').read()).group(1))"
The GraalVM version should match mx.trufflesqueak/suite.py.
Setup
Prefer the following manual setup below for local development:
# Install dependencies (Ubuntu 24.04)
sudo apt-get update
sudo apt-get install -y build-essential zlib1g-dev cmake python3-pip
# Install Oracle GraalVM matching the current suite version.
# If SDKMAN is available, this is the easiest option:
sdk install java <graalvm-version>-graal
sdk use java <graalvm-version>-graal
# Alternatively, use `mx fetch-jdk`.
# Install pre-commit hooks
python3 -m pip install --user pre-commit
pre-commit install
# Work from the parent directory of the TruffleSqueak checkout
export TS_WORKSPACE="$(cd .. && pwd)"
# Clone mx next to trufflesqueak and add it to PATH
git clone https://github.com/graalvm/mx.git "$TS_WORKSPACE/mx"
export PATH="$TS_WORKSPACE/mx:$PATH"
# Clone graal next to trufflesqueak and apply the checked-in patch
mx sforceimport
(cd "$TS_WORKSPACE/graal" && git apply "$TS_WORKSPACE/trufflesqueak/mx.trufflesqueak/graalvm-<graalvm-version>.patch")
Build and Run
Read mx.trufflesqueak/suite.py first. It is the quickest way to understand project names, imported suites, and distribution names.
# Build JVM standalone (faster to build) and find the target build directory
mx --env jvm build
mx --env jvm paths --output "TRUFFLESQUEAK_JVM_STANDALONE"
# Build native standalone (slower to build) and find the target build directory
mx --env native build --build-log important
mx --env native paths --output "TRUFFLESQUEAK_NATIVE_STANDALONE"
# Launch TruffleSqueak from the standalone's bin/ directory
trufflesqueak --headless --quiet -- --evaluate "<smalltalk code>"
# Run 10 iterations of the Towers benchmark from AWFY
trufflesqueak --headless --quiet -- --evaluate "AWFYHarness run: #('Towers' 10 1500)"
# Run TruffleSqueak tests
trufflesqueak --vm.ea --download-image trufflesqueak-test -- src/image-tests/runSqueakTests.st
# Run Cuis tests
trufflesqueak --vm.ea --download-image cuis-7.5-test -- -s src/image-tests/runCuisTests.st
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.
- 6d ago First seen · 95 lines · 1,043 tokens per session scan B 13f08973085d
trufflesqueak AGENTS.md is an instructions file published in the GitHub repository hpi-swa/trufflesqueak (319 stars, last pushed 4d ago), licensed MIT. It adds 1,043 tokens to every session, about $0.0052 per session on Opus 5. A static security scan graded it B with 1 finding (asks for root). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
plugins CLAUDE.md
Claude Code instructions for ihub-pub/plugins, covering project knowledge base, project structure, plugin id 速查表, version compatibility and commands(build / lint / test).
ethereumjs-monorepo AGENTS.md
AGENTS.md instructions for ethereumjs/ethereumjs-monorepo, covering agent notes, docs, cursor and habits.
postern CLAUDE.md
Claude Code instructions for punt-labs/postern, covering postern, what postern is, architecture, packages on disk and development model.
mocapi CLAUDE.md
Instructions for callibrity/mocapi, covering claude notes for mocapi, workflow, documentation layout, when a change requires an adr + design-doc update (rule) and architectural invariants.
smalltalk-interop-mcp-server CLAUDE.md
Claude Code instructions for mumez/smalltalk-interop-mcp-server, covering claude.md, project overview, development setup, environment variables and common commands.
smalltalk-validator-mcp-server CLAUDE.md
Claude Code instructions for mumez/smalltalk-validator-mcp-server, covering claude.md, project overview, architecture, validation tools and linting tools.