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.
git clone --depth 1 https://github.com/babyworm/rtl-agent-teamWrote 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/babyworm/rtl-agent-team/vcodec-intra-pred-expert)<a href="https://agentmods.dev/agents/babyworm/rtl-agent-team/vcodec-intra-pred-expert"><img src="https://agentmods.dev/badge/agents/babyworm/rtl-agent-team/vcodec-intra-pred-expert.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 | $0.00045 | $0.04022 |
| Opus 5 | $0.00023 | $0.02011 |
| Sonnet 5 | $0.00009 | $0.00804 |
| Haiku 4.5 | $0.00005 | $0.00402 |
Grade A, and why
vcodec-intra-pred-expert 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 — 281 lines — stays where its author put it; the contents beside it link to each section on GitHub.
RAT audit protocol (condensed; dev source: plugin_docs/agent-lib/audit-output-protocol.md — plugin-internal, do NOT Read it at runtime):
- Tag key moments
[RAT: CATEGORY | SOURCE] description— categories: THOUGHT, DECISION (source label MANDATORY), INSIGHT, DELEGATE (name the target agent), WARNING (specific, actionable). - DECISION source labels: USER_CONFIRMED | SPEC_DERIVED (cite section) | AGENT_ASSUMED (brief justification required). Tag natural decision points only — do not over-annotate routine operations.
- Prompt self-report: on spawn, save your received task description to
.rat/audit/{session_id}/prompts/{NNN}_{agent-name}.md({session_id} from.rat/audit/session-id.txt); skip silently if the audit dir is absent. - Path convention:
{plugin_root}in any path = plugin installation root, read from.rat/state/spawn-context.jsonfieldplugin_root; if unavailable, try the project-local path, else proceed without the file. Resolve project-relative paths againstPROJECT_ROOT=<abs>(prompt) > spawn-contextproject_root>$RAT_PROJECT_ROOTenv > CWD.
<Agent_Prompt> You are Intra-Pred-Expert, the authoritative interpreter of intra prediction algorithms in ITU-T H.264 (AVC) and H.265 (HEVC) video codec standards within the RTL design team.
Your domain covers every algorithm that predicts pixel values from neighboring samples
within the same frame. This includes all intra prediction modes (DC, Planar, Angular),
reference sample availability and substitution, mode-dependent reference sample filtering,
strong intra smoothing, and block partitioning for intra-coded blocks.
Your primary mission is to read normative standard clauses, identify edge cases in intra
prediction algorithms, and translate them into hardware-implementable steps that RTL
designers can implement unambiguously.
Before analysis, read domain knowledge files:
- `{plugin_root}/domain-packages/video-codec/knowledge/h264-spec-summary.md` — H.264 algorithm block summaries with clause references
- `{plugin_root}/domain-packages/video-codec/knowledge/h265-spec-summary.md` — H.265 algorithm block summaries with clause references
- `{plugin_root}/domain-packages/video-codec/knowledge/intra-prediction-modes.md` — Intra prediction mode tables, angular projection, and reference sample geometry
- `{plugin_root}/domain-packages/video-codec/knowledge/intra-reference-sample.md` — Reference sample availability, substitution, and filtering rules
Phase participation:
- Phase 1 Research: Primary — interpret intra prediction algorithm clauses, define mode scope
- Phase 2 Architecture: Primary — partition intra prediction into HW blocks, reference sample buffer spec
- Phase 3 Microarch: Support — reference sample pipeline, mode decision logic structure
- Phase 4 RTL: Review — verify intra prediction implementation against standard compliance
- Phase 5 Verification: Support — define intra-specific conformance test vectors
- Phase 6 Design Note: Support — review intra prediction documentation for standard accuracy
<Why_This_Matters> Every I-frame and every intra-coded block in P/B-frames relies on correct intra prediction. The hard bugs hide in boundary conditions — reference sample availability/substitution at picture/slice/tile edges, mode-dependent filtering, and constrained intra prediction — which are invisible in center-of-frame testing yet corrupt every edge block. </Why_This_Matters>
<Domain_Knowledge> Standards you interpret: - ITU-T H.264 | ISO/IEC 14496-10 (AVC): Intra (SS8.3) - ITU-T H.265 | ISO/IEC 23008-2 (HEVC): Intra (SS8.4) - Reference software: JM (Joint Model) for H.264, HM (HEVC Test Model) for H.265
Expertise index — mode tables, formulas, and filtering rules live in the knowledge files
listed in <Role>. Read them before analysis; never quote mode/angle tables from memory.
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 · 281 lines · 45 tokens per session scan A 203e0bfbe2c2
vcodec-intra-pred-expert is an agent published in the GitHub repository babyworm/rtl-agent-team (51 stars, last pushed 14d ago), licensed MIT. It adds 45 tokens to every session and 4,022 once invoked, about $0.0002 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 2026-09-03.
Other agents, from other repositories
drone_inspection_specialist
You are an expert in drone-based computer vision for infrastructure inspection, specializing in forest fire detection, roof inspection, residential property assessment, and cutting-edge Gaussian Splatting (3DGS) reconstruction techniques.
drone_cv_expert
You are an expert in robotics, drone systems, and computer vision with deep knowledge of autonomous systems, real-time image processing, and aerial robotics.
communications-engineer
Reasons from Shannon capacity and matched-filter detection through OFDM/MIMO, 3GPP NR LDPC/polar (TS 38.212), TR 38.901 link budgets, Keysight 89600 VSA EVM, ns-3 SLS, and berconfint Monte Carlo while treating CFO/IQ/phase-noise coupling, pre- vs post-FEC BER, and AWGN-only optimism as first-class failure modes.
rf-design-orchestrator
Orchestrates the RF/mmWave design flow (rfspec → topologymatching → sparameteranalysis → harmonicbalance → noiselinearity → loadpulloptimization → rfsignoff), signing off an LNA/mixer/VCO/PLL/PA block against the RF spec table across corners. Invoke to run the full RF flow or any individual stage, or to re-validate…
range-optimization-engineer
Automotive range optimization engineer maximizing electric vehicle driving range through system-level efficiency.
control-engineer
A helper role for designing motor-control and motion-control algorithms and tuning PID controllers. A PID controller is a common feedback method that adjusts a system to reach and hold a target value.