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 skills add m3taz-ahmed/ai-globals --skill reverse-engineering-lordgit clone --depth 1 https://github.com/m3taz-ahmed/ai-globalsWrote 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/skills/m3taz-ahmed/ai-globals/reverse-engineering-lord)<a href="https://agentmods.dev/skills/m3taz-ahmed/ai-globals/reverse-engineering-lord"><img src="https://agentmods.dev/badge/skills/m3taz-ahmed/ai-globals/reverse-engineering-lord/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/skills/m3taz-ahmed/ai-globals/reverse-engineering-lord"><img src="https://agentmods.dev/badge/skills/m3taz-ahmed/ai-globals/reverse-engineering-lord.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 | $0.00082 | $0.01470 |
| Opus 5.5 | $0.00033 | $0.00588 |
| Sonnet 5.5 | $0.00016 | $0.00294 |
| Haiku 4.5 | $0.00008 | $0.00147 |
Grade A, and why
reverse-engineering-lord 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 4d 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 — 62 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reverse Engineering Lord
[SKILL] reverse-engineering-lord
[OBJ] Explain, compare, decompile, or reconstruct shipped artifacts with REA (rea-agents CLI/MCP server). Compressed doctrine adapted from morluto/rea (MIT) .agents/skills/reverse-engineer-anything.
[RULES]
- [REQ] When to Use — Only when a claim depends on a shipped binary/package, decompilation, runtime evidence, or behavior NOT established by available source. Ordinary source-repo analysis → normal repo tools, never REA.
- [REQ] Connect Only When Needed — REA tools available → proceed directly; NEVER run doctor/setup before every investigation. Absent tools → diagnose the exact blocker (missing registration vs stale session vs missing provider) with
npx -y rea-agents@latest doctor --client <client> --json(clients: claude_code, codex, cursor, gemini_cli, windsurf, devin, vscode…), repair scoped, restart agent, resume task. - [REQ] Setup Discipline — Configuration repair starts with a read-only plan:
npx -y rea-agents@latest setup --client <client> --dry-run --json; show proposed paths + backups + skill replacement; apply--yesonly after approval.rea setupNEVER installs/upgrades Node, npm, Java, Ghidra, IDA, JADX, Kuna, Binwalk, Unblob — bring-your-own engines only. - [REQ] Route the Target First — Pick the first tool from the supplied target: extracted JS/Electron tree or ASAR →
analyze_javascript_application; archive members (ZIP/APK/IPA/DMG/PyInstaller) →open_binary(detail: "summary"first for large inventories); Apple plist/assets →open_bundle_resources; APK code/classes →inspect_android_package; .NET PE/CLI →inspect_managed_artifact; firmware →inspect_firmware_regions; open browser page →list_browser_targets. Runtime options are metadata, not files. - [REQ] Evidence Discipline — Distinguish observation / derivation / inference / unknown in every report. Preserve provenance (paths, digests, offsets, source locations). Unresolved identity or applicability stays UNKNOWN — never treated as absent or aligned. Keep complete inline Evidence when the server does not advertise retained references.
- [REQ] Read-Only Authority — Ghidra uses ephemeral DBs without modifying executable bytes; IDA analysis is read-only (never save/close an attached GUI DB); no mutation authority on Windows Ghidra P0. Fail closed: skip capabilities while scope/authority/side-effect questions are unresolved.
- [REQ] Local Checkout First — A full REA source checkout is installed OUTSIDE this repo: resolve
$REA_HOME, else sibling../reanext to the aizee root (tech-stack/useful-repos.mdis the registry). Invoke through the engine-aware wrappernode ../rea-cli.mjs <command> /abs/path --json(siblingrea-cli.mjsnext to the aizee root — it pins the checkout's.nvmrcNode underNVM_HOMEand injectsGHIDRA_INSTALL_DIR/JAVA_HOME); plainnode "$REA_HOME/scripts/rea.mjs"also works for engine-free tasks;npx -y [email protected]is the last resort. All return the same Evidence record, no MCP needed for static JS/.NET/artifact work. Native targets still require their selected engine (Hopper/Ghidra/IDA); a CLI fallback does NOT prove MCP registration. NEVER edit files inside the checkout; refresh it withgit -C <checkout> pull && npm ci && npm run build:cachedunder the pinned Node. Missingdist/→ runnpm ci && npm run build:cachedinside the checkout before use. - [REQ] Timeout Handling — Host timeout without a result ≠ work stopped; check provider startup deadline (
REA_GHIDRA_STARTUP_TIMEOUT_MS/REA_HOPPER_STARTUP_TIMEOUT_MSfor large binaries) before retrying an unchanged call. - [PROHIBIT] Never repeat an unchanged call after a timeout or assume an absent provider blocks unrelated targets — static JavaScript inspection needs neither Hopper nor Ghidra.
- [PROHIBIT] Never run target binaries, inject, mutate, or recover protections beyond declared scope — those need an explicit user scope decision first.
- [PROHIBIT] Never claim runtime reproduction you did not perform; label hypotheses and unverified coverage explicitly (
NOT_ENOUGH_INFObeats a confident guess). - [CMD] Providers — deep native: Hopper/Ghidra/IDA/Kuna/JEB + objdump/rizin/cutter/GDB; mobile: JADX/Apktool/adb; web: Chrome CDP/Playwright, HAR/SAZ captures; formats: WABT (wasm), tshark (pcap), sigrok (signals), binwalk/unblob (firmware), pwntools/GDB (ELF/coredumps). All caller-supplied.
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.
- 4d ago First seen · 62 lines · 82 tokens per session scan A 49ba62e6a56b
reverse-engineering-lord is a skill published in the GitHub repository m3taz-ahmed/ai-globals (5 stars, last pushed 5d ago), licensed MIT. It adds 82 tokens to every session and 1,470 once invoked, about $0.0003 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 2026-10-24.
Other skills, from other repositories
debug-optimize-lcp
Guides debugging and optimizing Largest Contentful Paint (LCP) using Chrome DevTools MCP tools. Use this skill whenever the user asks about LCP performance, slow page loads, Core Web Vitals optimization, or wants to understand why their page's main content takes too long to appear. Also use when the user mentions…
systematic-debugging
Use when debugging a failing test, build error, or runtime issue that isn't immediately obvious. Guides a 4-phase root cause analysis instead of random fix attempts.
log-error-digest
Analyze log files to troubleshoot errors, identify peak error periods, and produce error clustering, frequency statistics, and time distribution reports. Supports JSON, syslog, and Nginx formats with automatic detection. Use when a user uploads a .log file and asks to analyze errors, find patterns, debug issues, or…
cwv-optimizer
Diagnose and fix Core Web Vitals issues on AEM Edge Delivery Services pages. Goes deeper than generic CWV advice by understanding EDS-specific performance patterns including the 100KB LCP budget, E-L-D loading phases, block rendering behavior, and third-party script impact. Produces specific fixes for LCP, CLS, and…
incident-response-incident-response
Coordinate production incident triage, mitigation, investigation, recovery, and follow-up. Use when responding to outages or other incidents that require structured operational coordination.
ci-fixer
CI failures - read error, minimal fix, verify.