aistack-consensus

A human-approval checkpoint system for aistack agents. It lets an agent propose an action and pause until an authorised person approves or rejects it.

In plain words
What is it for?
Use it to create pending approvals, list or inspect them, and approve or reject proposed high-impact actions before they run.
Why use it?
It prevents unattended agents from carrying out risky or hard-to-reverse actions such as production deployments, database changes, mass edits or data deletion.

Skill for Claude CodeCodex

Part of the aistack plugin — 3 skills, 3 commands shipped together

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/blackms/aistack/consensus
Any agent
npx skills add blackms/aistack --skill consensus
Clone the repo
git clone --depth 1 https://github.com/blackms/aistack

Made for: Claude Code, Codex.

Or install aistack, the plugin that ships this one along with the rest of its 3 skills, 3 commands.

Per session 75 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 627 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 $0.00075 $0.00627
Opus 5 $0.00037 $0.00313
Sonnet 5 $0.00015 $0.00125
Haiku 4.5 $0.00007 $0.00063

Measured 3d ago against content hash 85533b20cdc1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

aistack-consensus 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.

plugin/skills/consensus/SKILL.md · 45 lines

How it starts

The opening of the file, as written. The whole thing — 45 lines — stays where its author put it; the contents beside it link to each section on GitHub.

aistack: Consensus checkpoints (human-in-the-loop)

aistack lets agents pause before high-impact actions and request explicit approval. This prevents autonomous agents from taking irreversible or risky steps without a human (or a designated approver) signing off.

When to use

  • Before an irreversible or high-blast-radius action: production deploy, DB schema migration, deleting data, force-push, mass file rewrite, spending money, sending external messages.
  • When the user wants approval gates, sign-off, or "ask me before doing X".
  • In an autonomous loop (see /aistack-pm) where some steps must not run unattended.

For low-risk, easily reversible actions, do NOT add a checkpoint — it just adds friction.

MCP tools (bundled aistack server)

Tool Purpose
consensus_check Propose an action and create a pending checkpoint requiring approval
consensus_list_pending List checkpoints awaiting a decision
consensus_get Inspect a specific checkpoint
consensus_approve Approve a pending checkpoint (the action may proceed)
consensus_reject Reject a checkpoint (the action must not proceed)

List exact schemas with npx @blackms/aistack mcp tools.

Workflow

  1. Detect the gate: classify the next action's risk. If it is irreversible / high-impact, do not execute it directly.
  2. Propose: call consensus_check describing the exact action, its blast radius, and how to undo it (if possible). This creates a pending checkpoint.
  3. Pause & surface: stop and tell the human what is pending and why, then wait. Do not proceed on assumption.
  4. Resolve: on consensus_approve, perform the action. On consensus_reject, abort cleanly, report, and propose a safer alternative if one exists.
  5. Record: store the decision and its rationale in aistack memory (see the aistack-memory skill) for an audit trail.

Tips

  • Make each proposal self-contained: a reviewer should be able to decide from the checkpoint text alone, without re-reading the whole session.
  • Default to a checkpoint when uncertain whether an action is reversible. Over-gating is annoying; under-gating is dangerous.
  • Combine with /aistack-review to run an adversarial pass before requesting approval, so the human reviews a vetted action.

Read the full file on GitHub · 45 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. 3d ago First seen · 45 lines · 75 tokens per session scan A 85533b20cdc1

Subscribe to this mod's changes

aistack-consensus is a skill published in the GitHub repository blackms/aistack (53 stars, last pushed 2mo ago), licensed MIT. It adds 75 tokens to every session and 627 once invoked, about $0.0004 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

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

brainstorming

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

obra/superpowers · 37 tokens

auto-perf-optimize

Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.

microsoft/vscode · 62 tokens

chat-perf

Run chat perf benchmarks and memory leak checks against the local dev build or any published VS Code version. Use when investigating chat rendering regressions, validating perf-sensitive changes to chat UI, or checking for memory leaks in the chat response pipeline.

microsoft/vscode · 51 tokens

chat-pet-sprite-creation

Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.

microsoft/vscode · 53 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens