Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Superstack Decide

ASecurity

A mid-work technical fork — "which way should we go", a design/architecture/tooling choice about to lock — or the user ruling on how the thing is built (conduct rulings are superstack-doctrine's). Also on "record this decision". Skip for statute-governed choices, forks needing building (superstack-spike), and trivially reversible picks (a log line).

2 stars
0 votes
0 copies
1 views
Added 9/19/2026
ai-agentsgo

Works with

cli

Security Analysis

A100/100

Scanned 9/19/2026

Install to Claude Code

$npx -y skills add debabsah/superstack --skill superstack-decide --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Superstack Decide?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Superstack Decide
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/debabsah-superstack-decide/badge)](https://www.skillsdirectory.com/skills/debabsah-superstack-decide)

More formats (shields.io, HTML) on the badges page.

Download with Pro
Files
SKILL.md
---
name: superstack-decide
description: A mid-work technical fork — "which way should we go", a design/architecture/tooling choice about to lock — or the user ruling on how the thing is built (conduct rulings are superstack-doctrine's). Also on "record this decision". Skip for statute-governed choices, forks needing building (superstack-spike), and trivially reversible picks (a log line).
---

# superstack-decide

**First act on every load:** append one line — `skill-load decide <date> <session-id>` — to `.superstack/receipts/loads.log` (create the dir if missing). Mechanical proof of routing; a fork that locked with no load line is countable.

A decision is real only when its rejected alternative and revisit trigger survive it. Chat history is not a record; six weeks later, "why did we choose X?" deserves a file, not an archaeology dig.

## The fork

1. **Frame it in one line** — what locks if we proceed, and how expensive reversal is (the reversibility class decides how much thinking this earns — R7).
2. **Price 2–4 options.** Each carries its downstream cost in one clause; one is marked recommended with the rejected cost stated (R6's shape: one question, a recommendation, the rejected cost). A menu never boxes the conclusion in — name what's outside it.
3. **Who decides:** reversible and in scope → decide now and record why. A one-way door, a genuine ambiguity, or anything touching an anchor → the human (R6). When the human's ruling is about *conduct* ("never do X again"), it's a statute — route to superstack-doctrine; a *design* choice lands here. One event can mint both.

## The record

- **House style wins.** If the repo already keeps decision records — `docs/adr/`, `ADR-*.md`, any existing convention — use *their* path and format; this shape fills silence, never fights a convention.
- **The first record in a silent repo is offered, never silent.** Creating `docs/decisions/` mints a new committed convention in someone's repo — say so and ask ("this locks an approach; want it recorded in docs/decisions/?"), create the directory on yes. If the owner declines, decisions live in the overlay's Conventions instead — durable on this machine only, and say that limit out loud.

On yes, create `docs/decisions/NNNN-<slug>.md` (zero-padded, first record `0001`; next = ls the directory):

```
# NNNN — <choice, imperative>
date: <YYYY-MM-DD> · decided by: <human|model — context>
context: <2–3 lines: the fork and what forced it>
options: <A — cost> / <B — cost> / …
chose: <X> because <Z>
revisit if: <the observation that reopens this>
supersedes: <NNNN — omit unless replacing one>
```

Same grammar as the task file's decision log (`chose X over Y because Z; revisit if W`) — this is that line's durable, committed form. Append-only: supersede, never edit (doctrine's rule, same reason) — and when a record is superseded, append one line to the OLD file, `superseded-by: NNNN`. An append, not an edit: the dead end carries a sign pointing at the live decision.

- **Committed on purpose:** the overlay is per-machine and destructively compacted; a decision record must survive machines, teammates, and compaction.
- **Promotion seam:** when superstack-ship retires a task file, decision-log lines that outlived the task promote here (ship's step 4 names this target).
- **Not every choice earns a file:** the test is the future question — if nobody will ask "why", the task-file line was enough. Records that can't name their revisit trigger are wishes, not decisions.

Attribution

debabsahdebabsah
View sourceMore from debabsah →
SSkills DirectorySkills Directory

Know which skills are safe — weekly.

Best new skills + every skill we flagged as malicious. From the team that scanned 103,619.

Join free

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Know which skills are safe — weekly.

Best new skills + every skill we flagged as malicious. From the team that scanned 103,619.

Join free

Related Skills

Caveman

Ultra-compressed communication mode that cuts output tokens while keeping technical accuracy. Levels: lite, full, ultra and the wenyan variants. Use for /caveman, "caveman mode", "talk like caveman", "be brief" or "less tokens".

1074701 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

693161 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3351 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

691 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants — handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents →