medsci-agent: Skill for OpenCode

.opencode/skills/operational-guardrails/SKILL.md

operational-guardrails is a skill for OpenCode from omar-A-hassan/medsci-agent. It costs 27 tokens per session (761 once invoked), scanned A, original, MIT.

A set of operating rules for MedSci agents, covering planning, tool use, retries, and evidence standards.

In plain words
What is it for?
It is for planning tasks before tool calls, running tool calls in sequence, applying retry limits, and checking that conclusions have adequate evidence.
Why use it?
It makes agent work more consistent and limits unsafe or poorly supported results.

Skill for OpenCode

Written for OpenCode: installed under .opencode/.

This is omar-A-hassan/medsci-agent's own configuration. It tells OpenCode how to work on medsci-agent itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything medsci-agent configures →

Reuse

Borrowing it

Nothing to install: this file belongs to omar-A-hassan/medsci-agent. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/omar-A-hassan/medsci-agent/main/.opencode/skills/operational-guardrails/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/omar-A-hassan/medsci-agent

Made for: OpenCode.

Wrote 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.

agentmods badge for operational-guardrails

README.md
[![agentmods](https://agentmods.dev/badge/skills/omar-a-hassan/medsci-agent/operational-guardrails/github.svg)](https://agentmods.dev/skills/omar-a-hassan/medsci-agent/operational-guardrails)
Your own site
<a href="https://agentmods.dev/skills/omar-a-hassan/medsci-agent/operational-guardrails"><img src="https://agentmods.dev/badge/skills/omar-a-hassan/medsci-agent/operational-guardrails/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.

agentmods 80×15 button for operational-guardrails

Your own site · 80×15
<a href="https://agentmods.dev/skills/omar-a-hassan/medsci-agent/operational-guardrails"><img src="https://agentmods.dev/badge/skills/omar-a-hassan/medsci-agent/operational-guardrails.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 27 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 761 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00027 $0.00761
Opus 5 $0.00014 $0.00380
Sonnet 5 $0.00005 $0.00152
Haiku 4.5 $0.00003 $0.00076

Measured 11d ago against content hash 7752562934db, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

operational-guardrails 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 11d 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.

.opencode/skills/operational-guardrails/SKILL.md · 59 lines

How it starts

The opening of the file, as written. The whole thing — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Operational Guardrails

These rules apply to every MedSci agent session. They override any conflicting behavior.

1. Planning Phase (Before Any Tool Call)

Before your first tool call, produce a brief plan:

  1. Classify the task type (lookup, analysis, synthesis, code execution).
  2. Identify which toolchains are needed and in what order.
  3. List data dependencies between steps.
  4. State stop conditions — what constitutes "done" and what would trigger an abort.
  5. Estimate evidence depth — is one tool call enough, or is multi-step synthesis required?

Only then begin executing. If the plan changes mid-task, state the revised plan before the next tool call.

2. Sequential Execution (No Exceptions)

Execute all MCP tool calls one at a time. Wait for each result before calling the next tool.

Why: MedGemma and TxGemma run locally inside MCP tools. Parallel calls queue on the same local model, causing MCP timeouts (error -32001).

Prohibited language in plans: Never write "I can run these in parallel", "these steps are independent", "simultaneously", or "at the same time" for tool calls. Every step executes sequentially — perceived independence is irrelevant. Write your plan as a numbered sequence and execute step 1 immediately after the plan.

3. Retry and Stop Policy

  • Per-tool retry limit: 1 retry on transient failure (timeout, rate limit). If it fails twice, skip it.
  • Session retry limit: No more than 3 total retries across all tools in one session.
  • Continue with partial data when a non-critical tool fails. Note the gap explicitly.
  • Halt and ask the user when a critical-path tool fails twice (e.g., the only data source for the query).
  • Never retry on permanent errors: CLI_UNAVAILABLE, ARTIFACT_PATH_FORBIDDEN, MODEL_NOT_FOUND.

4. Model Interpretation Fallback

When a tool returns model_used: false (MedGemma/TxGemma unavailable):

  • Return the raw data — it still has value.
  • Provide your own interpretation clearly labeled as "LLM interpretation (not domain model)."
  • Do not silently omit the raw data in favor of your interpretation.

Read the full file on GitHub · 59 lines

Changes

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.

  1. 11d ago First seen · 59 lines · 27 tokens per session scan A 7752562934db

Subscribe to this mod's changes

operational-guardrails is a skill published in the GitHub repository omar-A-hassan/medsci-agent (18 stars, last pushed 4d ago), licensed MIT. It adds 27 tokens to every session and 761 once invoked, about $0.0001 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.

Related

Other skills, from other repositories

llm-integration

LLM integration patterns for function calling, streaming responses, local inference with Ollama, and fine-tuning customization. Use when implementing tool use, SSE streaming, local model deployment, LoRA/QLoRA fine-tuning, or multi-provider LLM APIs.

yonatangross/orchestkit · 58 tokens

paseo-plugin

Build and manage trusted local Paseo plugins. Use when the user asks to create, edit, install, reload, enable, disable, remove, or troubleshoot a Paseo plugin; add lifecycle hooks; transform agent configuration, environment, MCP servers, or workspace creation; automate permissions or turn follow-ups; add a native…

getpaseo/paseo · 119 tokens

paseo

Paseo reference for managing projects, workspaces, workspace scripts, agents, schedules, and heartbeats.

getpaseo/paseo · 26 tokens

paseo-help

Answer questions about the Paseo product and app, including setup, configuration, connectivity, providers, workspaces, updates, logs, and troubleshooting. Use when a user inside Paseo asks how Paseo works, how to configure it, or why something is broken; use the paseo skill instead to operate agents and workspaces…

getpaseo/paseo · 75 tokens

paseo-advisor

Spin up a single agent as an advisor — second opinion on the current task. Use when the user says "advisor", "second opinion", "what does X think", or wants an outside take without delegating the work itself.

getpaseo/paseo · 52 tokens

paseo-handoff

Hand off the current task to another agent with full context. Use when the user says "handoff", "hand off", "hand this to", or wants to pass work to another agent.

getpaseo/paseo · 45 tokens