message-approval-policy

A policy for checking and approving messages before they are sent to people through email, chat, social media, SMS, or other channels.

In plain words
What is it for?
Use it when preparing, reviewing, sending, or investigating the status of an external-visible message.
Why use it?
It reduces the risk of sending a message to the wrong audience, sending unapproved content, exposing sensitive information, or falsely claiming delivery.

Skill for Claude CodeCodex

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 skills/codeinfinity1/stram/message-approval-policy
Any agent
npx skills add CodeInfinity1/Stram --skill message-approval-policy
Clone the repo
git clone --depth 1 https://github.com/CodeInfinity1/Stram

Made for: Claude Code, Codex.

Per session 30 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 610 The whole file, excluding the scripts and references it only reads on demand.
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 $0.00030 $0.00610
Opus 5 $0.00015 $0.00305
Sonnet 5 $0.00006 $0.00122
Haiku 4.5 $0.00003 $0.00061

Measured 2d ago against content hash 0a403712cc68, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

message-approval-policy 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 2d 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.

skills/communications/message-approval-policy/SKILL.md · 74 lines

How it starts

The opening of the file, as written. The whole thing — 74 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Message Approval Policy

Purpose

Keep the assistant trustworthy when it can speak through real channels. This skill defines the approval and verification discipline for messages, emails, social posts, SMS, chat replies, voice calls, and other external-visible communication.

When To Use

Use before sending any message to another person, room, phone number, work channel, social platform, email recipient, or call participant. Also use when reviewing prepared outbox messages or diagnosing whether something was sent.

Inputs And Evidence

  • Exact recipient or conversation ID.
  • Exact message text and media attachments.
  • Channel manifest, setup status, and approval state.
  • Reason for sending and supporting evidence.
  • Tool result status: prepared, blocked, sent, failed, or dry run.

Tool Map

  • channel_manifest
  • channel_setup_status
  • channel_message_prepare
  • channel_message_send
  • channel_outbox
  • cognitive_interaction_review
  • tool_describe
  • write_note

Workflow

  1. Classify the action as draft, prepare, send, post, speak, or call.
  2. Confirm recipient, audience, and message text.
  3. Check whether the channel has native direct send support or only prepared outbox support.
  4. Redact secrets, private logs, personal data, and unsupported claims.
  5. Prepare first when review is needed or credentials/runtime are missing.
  6. Send only through an approval-gated native tool and only after explicit approval.
  7. Report the exact final status and never blur "prepared" into "sent".

Native Implementation Boundaries

  • Approval policy is enforced through Stram risk levels, approval queue, channel tools, workflow tools, and audit logs.
  • Do not use upstream AgentGate/Passport/external reference policy code directly.
  • If a new external action tool is added, make it Stram-owned, schema-bound, and approval-gated.

Safety And Approval

  • The user owns final approval for external-visible communication.
  • High-risk content includes money, legal, medical, hiring, firing, secrets, reputation-sensitive posts, purchases, and irreversible public messages.
  • Keep a durable audit trail for sent or prepared messages.

Read the full file on GitHub · 74 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. 2d ago First seen · 74 lines · 30 tokens per session scan A 0a403712cc68

Subscribe to this mod's changes

message-approval-policy is a skill published in the GitHub repository CodeInfinity1/Stram (10 stars, last pushed 22d ago), licensed MIT. It adds 30 tokens to every session and 610 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 2026-08-31.