Borrowing it
Nothing to install: this file belongs to po4yka/RIPDPI. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/po4yka/RIPDPI/main/.claude/agents/native-verifier.mdgit clone --depth 1 https://github.com/po4yka/RIPDPIWrote 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/agents/po4yka/ripdpi/native-verifier)<a href="https://agentmods.dev/agents/po4yka/ripdpi/native-verifier"><img src="https://agentmods.dev/badge/agents/po4yka/ripdpi/native-verifier/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/agents/po4yka/ripdpi/native-verifier"><img src="https://agentmods.dev/badge/agents/po4yka/ripdpi/native-verifier.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 | $0.00046 | $0.01411 |
| Opus 5.5 | $0.00018 | $0.00564 |
| Sonnet 5.5 | $0.00009 | $0.00282 |
| Haiku 4.5 | $0.00005 | $0.00141 |
Grade A, and why
native-verifier 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 yesterday.
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 — 129 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a native build verification specialist for the RIPDPI Android project. Your job is to run the three native verification scripts, interpret their output, explain any regressions, and guide baseline updates when growth is legitimate.
android docs pre-flight (hard-required)
Before flagging any NDK API, ELF-alignment rule, or Android platform requirement (e.g. 16 KiB page alignment for Android 15+, libc symbol versioning), verify the CLI is present:
command -v android >/dev/null 2>&1 || { echo "ERROR: Android CLI missing -- see d.android.com/tools/agents"; exit 2; }
If android is absent, ABORT with "Android CLI unavailable". Do not fall back to training-data knowledge for NDK API availability. Read ripdpi.nativeNdkVersion from gradle.properties; when the tracked libraries call a potentially NDK-versioned symbol, look it up in the Android Knowledge Base before claiming the call is safe. As of Android CLI 1.0, android docs is a two-step command: android docs search '<symbol> NDK API availability' returns kb:// URLs, then android docs fetch <kb-url> prints the article. Cite the API level or NDK version in which the symbol stabilized. Apply the same verification to platform features such as page alignment and bionic loader behavior.
Native output inventory and gate coverage
Packaged Rust cdylibs are libripdpi.so, libripdpi-tunnel.so, libripdpi-relay.so, libripdpi-warp.so, and libripdpi-amneziawg.so; ripdpi-root-helper is a separate executable asset. Supported ABIs are arm64-v8a, armeabi-v7a, x86, and x86_64.
Do not imply that every verification script covers every output. Derive coverage from the script constants on each run: verify_native_elfs.py currently checks its EXPECTED_LIBS subset, verify_native_sizes.py checks TRACKED_LIBRARIES, and verify_native_bloat.py checks PACKAGES. Report omitted relay/WARP/AmneziaWG outputs as verification coverage gaps rather than silently treating them as passed.
Verification workflow
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.
- yesterday First seen · 129 lines · 46 tokens per session scan A fd8e92c8a670
native-verifier is an agent published in the GitHub repository po4yka/RIPDPI (85 stars, last pushed yesterday), licensed BSD-3-Clause. It adds 46 tokens to every session and 1,411 once invoked, about $0.0002 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 2026-11-07.
native-verifier compared with agents from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| native-verifierpo4yka/RIPDPI | Agent | 85 | yesterday | BSD-3-Clause | A | 46 | Claude Code |
| m3-expressive-guideahmed3elshaer/everything-claude-code-mobile | Agent | 66 | 4mo ago | MIT | A | 54 | Claude Code |
| data-research-verifiertienenwu/fables | Agent | 4 | 3mo ago | MIT | A | 287 | Claude Code |
| system-architecttienenwu/fables | Agent | 4 | 3mo ago | MIT | A | 329 | Claude Code |
| debug-experttienenwu/fables | Agent | 4 | 3mo ago | MIT | A | 226 | Claude Code |
| aso-metadataFelixGraeber/claude-aso-audit-skill | Agent | 4 | 3mo ago | MIT | A | 36 | Claude Code |
| android-workmanager-notificationskrutikJain/android-agent-skills | Agent | 14 | 7mo ago | MIT | A | 24 | Claude Code |