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 RedHatProductSecurity/prodsec-skills --skill bidirectional-filteringgit clone --depth 1 https://github.com/RedHatProductSecurity/prodsec-skillsWrote 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/redhatproductsecurity/prodsec-skills/bidirectional-filtering)<a href="https://agentmods.dev/skills/redhatproductsecurity/prodsec-skills/bidirectional-filtering"><img src="https://agentmods.dev/badge/skills/redhatproductsecurity/prodsec-skills/bidirectional-filtering/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/redhatproductsecurity/prodsec-skills/bidirectional-filtering"><img src="https://agentmods.dev/badge/skills/redhatproductsecurity/prodsec-skills/bidirectional-filtering.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 | $0.00045 | $0.00595 |
| Opus 5 | $0.00023 | $0.00298 |
| Sonnet 5 | $0.00009 | $0.00119 |
| Haiku 4.5 | $0.00005 | $0.00060 |
Grade A, and why
bidirectional-filtering 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 9d 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 — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Bidirectional Filtering with Runtime Guardrails
Security Requirement
A guardrails component SHOULD be deployed between the users/applications (or API gateway) and the models. This component acts as a gateway or proxy that inspects and acts on data flowing in both directions.
This skill refers to runtime guardrails (a deployed component), not model-level safety training.
Input Direction (User/App → Model)
Incoming prompts are raw or "tainted" input. The guardrails component analyzes them and applies rule-based actions:
| Action | Description |
|---|---|
| Block | Discard the prompt entirely, preventing it from reaching the model |
| Mask | Redact or obfuscate sensitive data (PII, credentials) before forwarding |
| Modify | Rewrite the prompt to remove dangerous patterns while preserving intent |
| Pass | Allow the prompt through unchanged |
Objectives:
- Prevent specific sensitive data from reaching the model
- Reduce the probability of prompt injection
- Enforce content policies on inputs
Output Direction (Model → User/App)
Model responses are inspected before delivery to the user or application:
| Action | Description |
|---|---|
| Block | Suppress the response if it contains harmful or policy-violating content |
| Mask | Redact sensitive data the model may have included in its response |
| Modify | Remove or rewrite problematic portions of the response |
| Pass | Deliver the response unchanged |
Objectives:
- Prevent leakage of sensitive training data
- Enforce content safety policies on outputs
- Filter harmful, biased, or off-topic responses
Architecture Position
User/App → API Gateway → Guardrails → Inference Engine → Model
↕ (inspects both directions)
User/App ← API Gateway ← Guardrails ← Inference Engine ← Model
Implementation Checklist
- Deploy a guardrails component between the API gateway and the inference engine
- Configure input rules for prompt analysis (block, mask, modify, pass)
- Configure output rules for response analysis (block, mask, modify, pass)
- Define PII detection and masking rules for both directions
- Define prompt injection detection rules for the input direction
- Define content safety policies for the output direction
- Log all guardrail actions (blocks, masks, modifications) for audit
- Monitor guardrail effectiveness and tune rules based on observed patterns
- Ensure the guardrails component does not become a single point of failure (deploy with redundancy)
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.
- 9d ago First seen · 67 lines · 45 tokens per session scan A 6d41a0782027
bidirectional-filtering is a skill published in the GitHub repository RedHatProductSecurity/prodsec-skills (52 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 45 tokens to every session and 595 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 2026-08-30.
Other skills, from other repositories
agent-platform-rag-engine-management
Manage and query Agent Platform RAG Engine Corpora and retrieve grounded contexts using the Google GenAI SDK. Use when listing RAG corpora or files, inspecting a corpus, retrieving contexts, or generating content grounded in a RAG corpus. Do not use for standard database queries (use SQL/Spanner skills), Google…
agent-platform-model-registry
Agent Platform Model Registry Management. Use when you need to upload, list, describe, update, or delete machine learning models (and their versions) in the Agent Platform Model Registry. Don't use for model training, model deployment to endpoints, or managing non-Agent Platform models.
foundry-config-setup
Resolve missing setup caused by a hardcoded Foundry project endpoint or model in a sample. Use when a sample fails because it uses a placeholder/hardcoded projectendpoint (for example "https://your-project.services.ai.azure.com") or a hardcoded model instead of reading them from the environment.
google-cloud-solution-agentic-analytics-spark-knowledge-catalog
Discovers requirements and generates guidance to design and deploy a governed, secure agentic-analytics solution for data that's distributed across Google Cloud, other cloud providers, or on-premises. Data that's outside Google Cloud (such as data from Databricks, Snowflake, Salesforce, SAP, or Oracle systems) is…
training-check
Interactively monitor training metrics from the current Codex session, periodically checking WandB or fallback logs for NaN, divergence, plateaus, and broken runs.
nemo-automodel-launcher-config
Configure NeMo AutoModel job launches for interactive runs, Slurm clusters, and SkyPilot cloud execution.