ironcurtain session.instructions.md

A set of review rules for the session layer, the part of an application that manages an AI conversation and its handoffs to other processes.

In plain words
What is it for?
Use it when reviewing session code, AI SDK calls, tool results, state transitions, escalation responses, and failed-message cleanup.
Why use it?
It prevents changes from breaking conversation state, tool-call handling, message recovery, or the file-based communication used for escalations.

Instructions file for GitHub Copilot

Install

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.

agentmods
npx agentmods add instructions/provos/ironcurtain/session
Clone the repo
git clone --depth 1 https://github.com/provos/ironcurtain

Made for: GitHub Copilot.

Per session 366 This file is loaded in full into every session.
When invoked 366 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
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 $0.00366 $0.00366
Opus 5 $0.00183 $0.00183
Sonnet 5 $0.00073 $0.00073
Haiku 4.5 $0.00037 $0.00037

Measured 3d ago against content hash 7f2b211988a0, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ironcurtain session.instructions.md 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 3d 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.

.github/instructions/session.instructions.md · 21 lines

What it actually says

Session Review Rules

The session layer manages the agent conversation lifecycle, including the AI SDK integration and escalation IPC.

Mandatory Checks

  • generateText() must use stopWhen: [stepCountIs(n)], NOT maxSteps. The maxSteps API does not exist in AI SDK v6.
  • tool() must use inputSchema, NOT parameters. The parameters API does not exist in AI SDK v6.
  • Tool call results use toolCalls[].input, NOT toolCalls[].args.
  • The tool type is ToolSet, NOT CoreTool.
  • SessionStatus follows a linear state machine: initializing -> ready -> processing -> ready -> closed. The processing state prevents reentrant sendMessage() calls. Do not remove or bypass this guard.
  • Escalation resolution writes a JSON file to the escalation directory. The file naming convention (response-{escalationId}.json) must match what the proxy expects. Changing the naming breaks the IPC contract.
  • writeUserContext() writes the user's message for the auto-approver and must use try/catch with silent failure. The escalation directory may not exist yet.
  • On sendMessage() failure, the message history must be truncated back to messageCountBefore to avoid leaving partial state. Removing this truncation corrupts the conversation history.
  • User-facing output MUST use process.stdout.write() or process.stderr.write(), NOT console.log(). The logger module intercepts console methods and redirects to the session log file.
  • The SandboxFactory type exists for test injection. The default factory creates a real sandbox; tests provide a mock. Do not remove this abstraction.
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. 3d ago First seen · 21 lines · 366 tokens per session scan A 7f2b211988a0

Subscribe to this mod's changes

ironcurtain session.instructions.md is an instructions file published in the GitHub repository provos/ironcurtain (596 stars, last pushed 3d ago), licensed Apache-2.0. It adds 366 tokens to every session, about $0.0018 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 instructions, from other repositories

agentos CLAUDE.md

Claude Code instructions for rivet-dev/agentos, covering agentos, boundaries, security model, sqlite schema ownership and runtime and registry.

rivet-dev/agentos · 4,057 tokens

agent-sandbox AGENTS.md

Instructions for kubernetes-sigs/agent-sandbox, covering agents.md, project summary, repository layout, agent skills and build, test, lint.

kubernetes-sigs/agent-sandbox · 4,218 tokens

agent-sandbox copilot-instructions.md

Instructions for kubernetes-sigs/agent-sandbox: Project Context & Architecture: Refer to AGENTS.md for full project background, module layout, toolchain versions, and core conventions.

kubernetes-sigs/agent-sandbox · 293 tokens

mcp-server-starrocks CLAUDE.md

Instructions for StarRocks/mcp-server-starrocks, covering claude.md, project overview, development commands, run the server directly for testing and run with test mode to verify table overview functionality.

StarRocks/mcp-server-starrocks · 907 tokens

after-effects-mcp AGENTS.md

Instructions for JUNKDOGE-JOE/after-effects-mcp, covering repository development and delivery rules, 0.0 read the current architecture direction first, 0. user authorization is the scope boundary, 1. measure outcomes, not activity and 2. prioritize by dependency and user value.

JUNKDOGE-JOE/after-effects-mcp · 6,738 tokens

Symbiont AGENTS.md

Instructions for ThirdKeyAI/Symbiont, covering symbiont — agent instructions, project structure, build and test, code style and commit guidelines.

ThirdKeyAI/Symbiont · 4,039 tokens