Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/marcosd4h/DeepExtractRuntimenpx agentmods add commands/marcosd4h/deepextractruntime/reconstruct-typesWrote 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/marcosd4h/deepextractruntime/reconstruct-types)<a href="https://agentmods.dev/commands/marcosd4h/deepextractruntime/reconstruct-types"><img src="https://agentmods.dev/badge/commands/marcosd4h/deepextractruntime/reconstruct-types/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/marcosd4h/deepextractruntime/reconstruct-types"><img src="https://agentmods.dev/badge/commands/marcosd4h/deepextractruntime/reconstruct-types.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 | $0.00000 | $0.01255 |
| Opus 5.5 | $0.00000 | $0.00502 |
| Sonnet 5.5 | $0.00000 | $0.00251 |
| Haiku 4.5 | $0.00000 | $0.00126 |
Grade A, and why
reconstruct-types 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 — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reconstruct Types
Overview
Reconstruct C/C++ struct and class definitions from a module's decompiled code by scanning memory access patterns, vtable contexts, and mangled names across all functions. Generates compilable C++ header files with per-field confidence annotations.
The text after /reconstruct-types specifies a module and optionally a class name:
/reconstruct-types appinfo.dll-- reconstruct all types in the module/reconstruct-types appinfo.dll CSecurityDescriptor-- reconstruct a specific class/reconstruct-types appinfo.dll --include-com-- include COM interface reconstruction/reconstruct-types appinfo.dll --validate-- validate generated header against assembly
IMPORTANT: Execution Model
This is an execute-immediately command. Do NOT present anything for user confirmation. Run and write the final reconstructed types straight to the chat as your response. The user expects to see the completed output.
Workspace Protocol
This command uses the type-reconstructor agent's reconstruct_all.py pipeline, plus optional validation:
- Create
.claude/workspace/<module>_reconstruct_types_<timestamp>/. - Pass
--workspace-dir <run_dir>and--workspace-step reconstructto thereconstruct_all.pyinvocation. - Keep only summary output in context; read full intermediate data from
<run_dir>/reconstruct/results.jsonwhen presenting results. - Use
<run_dir>/manifest.jsonto track completed/failed steps.
Execution Context
IMPORTANT: Any inline Python that imports
helpers.*must run withcd <workspace>/.claude(so the.claude/directory is onsys.path), not from the workspace root. Script invocations likepython .claude/skills/.../script.pycan be run from the workspace root because those scripts manage their own path setup.
Steps
Step 0: Preflight Validation
Validate arguments using helpers.command_validation.validate_command_args("reconstruct-types", {"module": "<module>"}).
If validation fails, report the errors and stop. On success, use result.resolved["db_path"] for subsequent script calls.
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 · 103 lines · 0 tokens per session scan A 3d954d2c2223
reconstruct-types is a command published in the GitHub repository marcosd4h/DeepExtractRuntime (21 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,255 tokens. 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 commands, from other repositories
cpp-test
Enforce TDD workflow for C++. Write GoogleTest tests first, then implement. Verify coverage with gcov/lcov.
verify-claim
A command for checking whether a C++ claim is true by compiling and running a small example, with reference material when needed.
review
Review a change against the rule-driven modern C++ contract with bounded discovery.
cpp-build
Fix C++ build errors, CMake issues, and linker problems incrementally. Invokes the cpp-build-resolver agent for minimal, surgical fixes.
cpp
C++ coding standards and best practices.
unreal
Create actors, set up the Gameplay Framework, build abilities, or debug an Unreal Engine 5 project.