sdlc-loop

Routes the SDLC development loop by reporting feature status and the next command.

Updated Jun 1, 2026
One-click install
npx skills add https://github.com/juartaurus98/codebase --skill sdlc-loop-juartaurus98
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sdlc-loop
Source: https://github.com/juartaurus98/codebase/tree/main/.claude/skills/sdlc-loop
Command: npx skills add https://github.com/juartaurus98/codebase --skill sdlc-loop-juartaurus98

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? In a multi-phase spec-driven development workflow (PRD, SRS, tech docs, code, tests, review), it is hard to know where each feature stands, which single command to run next, and what is blocking progress. This Skill reads workflow state and answers "what now" with exactly one answer, instead of regenerating documents or code itself. ## Core Features & Use Cases - Status routing (default mode): Reads workflow state and reports each feature's phase, loop count, blockers, and the single next command, prioritizing blocked inputs, edited artifacts, open findings, and gate verdicts. - Blast radius analysis (--blast): Diffs scenario spec hashes against the trace file to show exactly which code, tests, prompts, and gates are impacted by a spec change before any regeneration happens. - Framework self-check (--doctor): Validates the framework itself against data-defined checks (forbidden literals, path reality, gate coverage, cache freshness) and reports pass/fail per file without auto-fixing. - Bug-fix entry point (--fix-bug): Enters the loop mid-chain for bugfixes, enforcing diagnosis-first classification, a failing regression test before any fix, and promotion to SRS when the spec is wrong. - Use Case: After editing a feature's spec file, run the blast mode to see that only UC1's scenarios changed, which implementation and test files are impacted, and which gates flip to stale — so you regenerate only the affected scope. ## Quick Start Ask the assistant to run /sdlc to report where every feature stands and which single command comes next in the development loop.

Frequently Asked Questions about sdlc-loop

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find out which command to run next in an SDLC workflow?▼

Run the status mode with no arguments. It reads the workflow state, lists each feature's phase and blockers, and outputs exactly one next command based on blocked inputs, edited artifacts, open findings, and gate verdicts.

How do I check the impact of a spec change before regenerating code?▼

Use the --blast mode with a feature ID. It diffs scenario spec hashes against the trace file and lists the changed scenarios, impacted code, tests, prompts, and gates that flip to stale, scoped to only the affected use cases.

How do I handle a bug in a spec-driven development loop?▼

Use --fix-bug with a bug ID and symptom. It creates a bugfix state entry, requires classifying the failure against loop routes, mandates a failing regression test before any fix, and promotes the issue to SRS if the spec was wrong.

Does this skill generate code or documents directly?▼

No. It is a router only — it reads state and findings, then names the single command to run next. All generation is handed off to other commands such as /techdoc, /srs, /unittest, or /gen-code.

What happens when the loop keeps iterating without progress?▼

An anti-thrash guardrail stops the loop after a configured number of iterations with no gate advancing. It writes a blocking open question, reports it, and issues no further generation commands until a human answers.