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/n0zom1z0/th08/agents-mdgit clone --depth 1 https://github.com/N0zoM1z0/th08Wrote 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/n0zom1z0/th08/agents-md)<a href="https://agentmods.dev/instructions/n0zom1z0/th08/agents-md"><img src="https://agentmods.dev/badge/instructions/n0zom1z0/th08/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 | $0.01225 | $0.01225 |
| Opus 5 | $0.00613 | $0.00613 |
| Sonnet 5 | $0.00245 | $0.00245 |
| Haiku 4.5 | $0.00122 | $0.00122 |
Grade A, and why
th08 AGENTS.md 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 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.
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 — 97 lines — stays where its author put it; the contents beside it link to each section on GitHub.
TH08 reconstruction agent rules
This repository reconstructs one exact binary: the original Japanese TH08
version 1.00d executable identified below and in reccmp-project.yml. Do not
analyze or substitute a localized, patched, trial, or earlier executable.
Before changing reconstruction state
- Read
docs/ARCHITECTURE.md,docs/RE_WORKFLOW.md, and the relevant source. - Verify
resources/th08.exe: SHA-256330fbdbf58a710829d65277b4f312cfbb38d5448b3df523e79350b879213d924and size840704bytes. - Confirm the address and symbol in
config/mapping.csvand the appropriateconfig/reccmp-*.csvfiles before editing. - Inspect target instructions with a target-safe backend. Reconcile function boundaries against disassembly; Ghidra/IDA function extents are evidence, not accepted compiler boundaries.
- Build and compare the smallest affected source/object before broad edits.
Evidence and status rules
- Keep target observations, upstream reconstruction claims, adjacent-version inferences, and unknowns distinct.
- GensokyoClub/th08 is the imported reconstruction baseline. Its complete Git history and original contributor attribution must remain intact.
- The N0zoM1z0/th07 reconstruction supplies this repository's workflow and structure. TH07 source and TH06 are supporting evidence only: confirm names, ABI, field offsets, side effects, and boundaries against TH08 1.00d.
- Never mechanically paste decompiler output as source.
config/implemented.csvselects upstream implementations; it is not proof of an exact match. Likewise, an entry inconfig/mapping.csvorconfig/reccmp-functions.csvis a mapping, not a matching claim.- Only a reproducible reccmp/objdiff result against the exact target may support an exact-match claim. Do not invent or manually transcribe progress numbers.
Analysis database safety
IDA MCP is attached to whichever file is open in the GUI and has no reliable
program selector. Treat all TH08 IDA queries as blocked until the GUI is
switched to TH08 and the database is attested under docs/IDA_MCP.md.
The canonical SHA-256 applies to resources/th08.exe; an IDB or its loader
metadata can report a different input/container hash. That difference is not
itself evidence of a different mapped image. In that case, verify the
canonical file first, then compare IDA bytes at .text start, entry point,
the function being edited, and at least two separated .text locations
against resources/th08.exe. Only byte-for-byte agreement plus matching image
base/entry/.text extent permits IDA semantic evidence. Exact-match claims
remain reproducible comparisons against the canonical file only.
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 · 97 lines · 1,225 tokens per session scan A b4d7c23673bd
th08 AGENTS.md is an instructions file published in the GitHub repository N0zoM1z0/th08 (158 stars, last pushed 3d ago), licensed MIT. It adds 1,225 tokens to every session, about $0.0061 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.
Other instructions, from other repositories
ida-pro-mcp CLAUDE.md
Instructions for mrexodia/ida-pro-mcp, covering claude.md, what this project is, core implementation rules, ida thread safety and api conventions.
sm64ds-decomp AGENTS.md
AGENTS.md instructions for tangosdev/sm64ds-decomp, covering posting a pr (for ai agents and their humans), the one rule that matters, what a change looks like, which directory under src/ and shared headers (include/).
dembrandt CLAUDE.md
Claude Code instructions for dembrandt/dembrandt, covering claude.md, system instruction, development commands, running the tool and testing.
Reversecore_MCP AGENTS.md
AGENTS.md instructions for sjkim1127/Reversecore_MCP, covering agent customizations for reversecoremcp, project overview, quick commands, development setup and create virtual environment.
kuna AGENTS.md
AGENTS.md instructions for Noelo-Lab/kuna, a project described as: An agent-first decompiler designed to be refined by other agents. Kuna is written in Rust and was originally ported from Ghidra.
windiff CLAUDE.md
Claude Code instructions for ergrelet/windiff, a project described as: Tool that allows comparing symbol, type and syscall information of Microsoft Windows binaries across different versions of the OS, using a Web UI and/or LLMs.