/sdlc-review
Review Kanban handoffs and route verified outcomes.
$ npx -y skills add NousResearch/hermes-agent --skill sdlc-review --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
- Slash command
/sdlc-review
Context preview
The summary Claude sees to decide when to auto-load this skill.
Review Kanban handoffs and route verified outcomes.
SKILL.md
sdlc-review.SKILL.mdname: sdlc-review
description: Review Kanban handoffs and route verified outcomes.
version: 1.1.0
author: Jakub Wolniewicz (@frizikk) + Hermes Agent
license: MIT
platforms: [linux, macos, windows]
metadata:
hermes:
tags: [kanban, review, quality, verification]
category: devops
requires_toolsets: [kanban]
environments:
- kanbanSDLC Review Skill
Independently verify work handed from a Kanban implementation run to the review lane, then approve it, request changes, or escalate. This skill reviews the deliverable and its evidence; it does not take over the implementer's work.
When to Use
Use this skill when all of the following are true:
- the dispatcher spawned you for a task claimed from the `review` lane;
- an implementer submitted a `review_requested` handoff;
- the task needs an independent verdict before it can be completed.
Do not use it for a separate downstream review card. A downstream card is ordinary implementation work with a review-oriented specification and completes through its own lifecycle.
Prerequisites
- A Kanban worker context with the current task and run identifiers.
- Native Kanban tools: `kanban_show`, `kanban_comment`, `kanban_complete`, `kanban_request_changes`, and `kanban_block`.
- Workspace access through `read_file`, `search_files`, and `terminal` when the deliverable is code.
- The task's original specification, acceptance criteria, handoff summary, and prior run history must be available through `kanban_show`.
How to Run
This skill is loaded automatically by the review dispatcher. Start with `kanban_show` before inspecting files or choosing a verdict.
1. Read the task specification and the latest `review_requested` handoff. 2. Inspect the actual deliverable and run relevant verification. 3. Choose exactly one verdict: approve, request changes, or escalate. 4. Record concrete evidence in the terminal Kanban transition.
Quick Reference
| Verdict | When | Final action | |---|---|---| | Approve | Acceptance criteria and verification pass | `kanban_complete` | | Request changes | Correctable implementation defects remain | `kanban_comment`, then `kanban_request_changes` | | Escalate | A human decision or external prerequisite is required | `kanban_block` |
A requested-changes transition returns the task to its original implementer. When that implementer requests review again without naming a reviewer, the persisted reviewer provenance routes the re-review back to the same reviewer profile.
Review Lenses
Vary how you look at the work on each round instead of repeating the same inspection. Decorrelated lenses catch different defect classes: a cold read of the artifact surfaces design and correctness problems that the implementer's narrative would have framed away, execution surfaces claims that do not reproduce, and a strict contract audit surfaces quiet scope drift. Repeating the round-1 lens on round 3 mostly re-finds what round 1 already found.
Determine the current round from the history the task record already gives you: count the `changes_requested` entries in the "Prior attempts on this task" section of your worker context (also visible as prior runs in `kanban_show`). The current review round is that count plus one. Round 1 therefore shows zero `changes_requested` attempts; round 2 shows one; and so on.
| Round | Lens | How to apply it | |---|---|---| | 1 | Artifact | Read the diff or deliverable cold, before the implementer's summary. Form an independent judgment, then compare it against the handoff narrative and investigate every mismatch. | | 2 | Execution | Check out the work and actually run it via `terminal`: build, test, and exercise the reported behavior yourself. Verify each handoff claim empirically instead of re-reading the artifact. | | 3+ | Contract | Re-read the ORIGINAL task body and acceptance criteria, then audit the deliverable strictly against them. Also verify that every item from every prior `kanban_request_changes` round actually landed. |
The baseline duties in the Procedure section still apply on every round; the lens sets which inspection you lead with and weight most heavily.
Lens variation for ad-hoc review fan-outs
The same principle applies outside the Kanban review lane. When spawning multiple parallel reviewers via `delegate_task`, give each reviewer a different lens — one diff-only brief, one full-context brief, one checkout-and-run brief — rather than identical briefs. Identical briefs produce correlated verdicts and duplicate findings; varied briefs cover more defect classes for the same review spend.
Procedure
1. Orient from the durable task record
Call `kanban_show` and identify:
- the original task body and acceptance criteria;
- the latest implementation summary and structured metadata;
- changed files, commit identifiers, and test evidence;
- comments and decisions from earlier runs;
- findings from prior review rounds.
Treat the handoff as a claim to verify, not as proof that the work is correct.
2. Compare requested behavior with delivered behavior
Map every acceptance criterion to concrete implementation or output evidence. Note omissions, changed semantics, and unrelated scope before deciding whether to run deeper checks.
For code work:
1. Use `read_file` and `search_files` to inspect the changed paths and their callers. 2. Use `terminal` to inspect the diff and run the project's existing focused tests, lint, type checks, or build commands. 3. Exercise the reported failure path and at least one ordinary control path when practical. 4. Check error handling, edge cases, concurrency boundaries, data preservation, security boundaries, and cross-platform behavior relevant to the change. 5. Confirm that tests assert behavior rather than merely snapshotting source text or constants.
For non-code work:
1. Inspect the complete deliverable rather than only its summary. 2. Check correctness, completeness, formatting, and provenance. 3. Vali
Read more
name: sdlc-review
description: Review Kanban handoffs and route verified outcomes.
version: 1.1.0
author: Jakub Wolniewicz (@frizikk) + Hermes Agent
license: MIT
platforms: [linux, macos, windows]
metadata:
hermes:
tags: [kanban, review, quality, verification]
category: devops
requires_toolsets: [kanban]
environments:
- kanbanSDLC Review Skill
Independently verify work handed from a Kanban implementation run to the review lane, then approve it, request changes, or escalate. This skill reviews the deliverable and its evidence; it does not take over the implementer's work.
When to Use
Use this skill when all of the following are true:
- the dispatcher spawned you for a task claimed from the `review` lane;
- an implementer submitted a `review_requested` handoff;
- the task needs an independent verdict before it can be completed.
Do not use it for a separate downstream review card. A downstream card is ordinary implementation work with a review-oriented specification and completes through its own lifecycle.
Prerequisites
- A Kanban worker context with the current task and run identifiers.
- Native Kanban tools: `kanban_show`, `kanban_comment`, `kanban_complete`, `kanban_request_changes`, and `kanban_block`.
- Workspace access through `read_file`, `search_files`, and `terminal` when the deliverable is code.
- The task's original specification, acceptance criteria, handoff summary, and prior run history must be available through `kanban_show`.
How to Run
This skill is loaded automatically by the review dispatcher. Start with `kanban_show` before inspecting files or choosing a verdict.
1. Read the task specification and the latest `review_requested` handoff. 2. Inspect the actual deliverable and run relevant verification. 3. Choose exactly one verdict: approve, request changes, or escalate. 4. Record concrete evidence in the terminal Kanban transition.
Quick Reference
| Verdict | When | Final action | |---|---|---| | Approve | Acceptance criteria and verification pass | `kanban_complete` | | Request changes | Correctable implementation defects remain | `kanban_comment`, then `kanban_request_changes` | | Escalate | A human decision or external prerequisite is required | `kanban_block` |
A requested-changes transition returns the task to its original implementer. When that implementer requests review again without naming a reviewer, the persisted reviewer provenance routes the re-review back to the same reviewer profile.
Review Lenses
Vary how you look at the work on each round instead of repeating the same inspection. Decorrelated lenses catch different defect classes: a cold read of the artifact surfaces design and correctness problems that the implementer's narrative would have framed away, execution surfaces claims that do not reproduce, and a strict contract audit surfaces quiet scope drift. Repeating the round-1 lens on round 3 mostly re-finds what round 1 already found.
Determine the current round from the history the task record already gives you: count the `changes_requested` entries in the "Prior attempts on this task" section of your worker context (also visible as prior runs in `kanban_show`). The current review round is that count plus one. Round 1 therefore shows zero `changes_requested` attempts; round 2 shows one; and so on.
| Round | Lens | How to apply it | |---|---|---| | 1 | Artifact | Read the diff or deliverable cold, before the implementer's summary. Form an independent judgment, then compare it against the handoff narrative and investigate every mismatch. | | 2 | Execution | Check out the work and actually run it via `terminal`: build, test, and exercise the reported behavior yourself. Verify each handoff claim empirically instead of re-reading the artifact. | | 3+ | Contract | Re-read the ORIGINAL task body and acceptance criteria, then audit the deliverable strictly against them. Also verify that every item from every prior `kanban_request_changes` round actually landed. |
The baseline duties in the Procedure section still apply on every round; the lens sets which inspection you lead with and weight most heavily.
Lens variation for ad-hoc review fan-outs
The same principle applies outside the Kanban review lane. When spawning multiple parallel reviewers via `delegate_task`, give each reviewer a different lens — one diff-only brief, one full-context brief, one checkout-and-run brief — rather than identical briefs. Identical briefs produce correlated verdicts and duplicate findings; varied briefs cover more defect classes for the same review spend.
Procedure
1. Orient from the durable task record
Call `kanban_show` and identify:
- the original task body and acceptance criteria;
- the latest implementation summary and structured metadata;
- changed files, commit identifiers, and test evidence;
- comments and decisions from earlier runs;
- findings from prior review rounds.
Treat the handoff as a claim to verify, not as proof that the work is correct.
2. Compare requested behavior with delivered behavior
Map every acceptance criterion to concrete implementation or output evidence. Note omissions, changed semantics, and unrelated scope before deciding whether to run deeper checks.
For code work:
1. Use `read_file` and `search_files` to inspect the changed paths and their callers. 2. Use `terminal` to inspect the diff and run the project's existing focused tests, lint, type checks, or build commands. 3. Exercise the reported failure path and at least one ordinary control path when practical. 4. Check error handling, edge cases, concurrency boundaries, data preservation, security boundaries, and cross-platform behavior relevant to the change. 5. Confirm that tests assert behavior rather than merely snapshotting source text or constants.
For non-code work:
1. Inspect the complete deliverable rather than only its summary. 2. Check correctness, completeness, formatting, and provenance. 3. Vali
The self-improving AI agent built by Nous Research. It's the only agent with a built-in learning loop — it creates skills from experience, improves them during use, nudges itself to persist knowledge, searches its own past conversations, and builds a
Repo: NousResearch/hermes-agent

