sdd-verify

sdd-verify is a command for coding agents from Gentleman-Programming/gentle-ai. It costs 9 tokens per session (550 once invoked), scanned A, original, MIT.

A verification command for checking whether an implementation matches its written proposal, design, specifications, and task list. SDD refers to a specification-driven development process in which those documents guide the implementation.

In plain words
What is it for?
Use it as the final independent verification step for an active SDD change, then follow the refreshed status to decide whether to review or archive it.
Why use it?
It helps detect requirements or runtime mismatches before the change is finalized.

Command

About the project

Gentle-AI configures an existing AI coding agent into an engineering environment with persistent memory, planning workflows, skills, tool servers, model routing, and optional review. Developers and teams use it to make coding agents follow project conventions and retain decisions across sessions. The catalogue entries are its skills, commands, agents, and instruction.

Gentleman-Programming/gentle-ai · 6,255 stars · on GitHub · gentle-ai.gentlemanprogramming.com

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 commands/gentleman-programming/gentle-ai/sdd-verify
Clone the repo
git clone --depth 1 https://github.com/Gentleman-Programming/gentle-ai

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 sdd-verify

README.md
[![agentmods](https://agentmods.dev/badge/commands/gentleman-programming/gentle-ai/sdd-verify.svg)](https://agentmods.dev/commands/gentleman-programming/gentle-ai/sdd-verify)
Your own site
<a href="https://agentmods.dev/commands/gentleman-programming/gentle-ai/sdd-verify"><img src="https://agentmods.dev/badge/commands/gentleman-programming/gentle-ai/sdd-verify.svg" alt="Measured on agentmods" height="20"></a>
Per session 9 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 550 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 5.1 $0.00009 $0.00550
Opus 5 $0.00005 $0.00275
Sonnet 5 $0.00002 $0.00110
Haiku 4.5 $0.00001 $0.00055

Measured 3d ago against content hash bacb5b75aad6, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

sdd-verify 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.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

internal/assets/opencode/commands/sdd-verify.md · 32 lines

What it actually says

You are the gentle-orchestrator, not an SDD executor. This command may launch the hidden sdd-verify sub-agent only after the orchestration gates below pass.

CONTEXT:

  • Working directory: before doing anything else, run git rev-parse --show-toplevel 2>/dev/null || pwd with your bash tool and use the returned path as the authoritative workspace. In OpenCode Desktop (Electron) the parse-time interpolation resolves to the app data directory, not the project.
  • Current project: the basename of the detected workspace above.

HARD GATES:

  1. SDD Session Preflight must already be complete for this session. It must include execution mode, artifact store, chained PR strategy, and review budget. If missing, ask the exact orchestrator preflight prompt and STOP. Do not run verify in the same turn.
  2. sdd-init must already exist or be run after preflight, per the orchestrator init guard.
  3. Resolve the active change using the status contract. If $ARGUMENTS is missing or ambiguous, ask the user to choose and STOP. Do not guess.
  4. Produce structured status before acting. Use the resolved artifact store from session preflight; do not hardcode Engram.
  5. The active change must have tasks and implementation evidence. Missing specs/design may be handled gracefully by the verify skill, but missing tasks means there is nothing to verify.
  6. actionContext must be safe for verification. If status reports workspace-planning, STOP and explain that full workspace implementation verification is not supported in this slice.

DEPENDENCY CHECK:

  • If tasks are missing, do NOT verify.
  • Tell the user what is missing and suggest /sdd-continue <change> or /sdd-apply <change> as appropriate.

TASK: If all gates pass, launch the hidden sdd-verify sub-agent with the structured status, available artifacts, and strict TDD instructions if sdd-init detected strict TDD — there is no review-state prerequisite. This is the single independent requirements/runtime verification; a contradiction escalates and never starts another review/refuter/fix loop. After verify returns, rerun native SDD status and route only from its refreshed nextRecommended. Present the post-verify review offer only if that refreshed status contains a fresh reviewOffer block; if it is absent, proceed toward archive with no review ceremony.

Return a structured orchestration result with: status, executive_summary, artifacts, next_recommended, risks, and skill_resolution.

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 Changed · -11 lines bacb5b75aad6
  2. 6d ago First seen · 43 lines · 9 tokens per session scan A 2f32af968c35

Subscribe to this mod's changes

sdd-verify is a command published in the GitHub repository Gentleman-Programming/gentle-ai (6,255 stars, last pushed today), licensed MIT. It adds 9 tokens to every session and 550 once invoked, about $0.0000 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.