All authors
Claude Skills by DarrenBenson
github.com/DarrenBenson7 skills0 installs7 views
- Sdlc StudioThe antidote to vibe coding: a full software engineering team at your fingertips. Create or reverse-engineer PRDs, TRDs, personas, epics, and user stories with acceptance criteria, then plan, implement, test, and verify code against them - plus an autonomous Goal-Driven sprint loop that drives a prioritised batch to a goal, an adversarial audit, status dashboards, reconciliation, change requests, RFCs, test specs, test automation, and bug tracking. Use when asked about PRDs, requirements docu...Votes: 0GitHub stars: 23
- Best Practices> **Machine-surfaced:** `scripts/disclosure.py` (advisory) checks much of this list - Load-when > markers, orphan-freedom, scripts `--help`/executable, template placeholders, SKILL.md sections. Based on [official Claude Code documentation](https://code.claude.com/docs/en/skills) and [Anthropic engineering guidance](https://www.anthropic.com/engineering/equipping-agents-for-the-real-world-with-agent-skills).Votes: 0GitHub stars: 23
- Audit ProfilesThe packaged lens pack for auditing an **agent skill itself**. The four lenses proven on the 2026-06-20 self-audit run. Load this pack as the profile when the project under audit *is* a skill (vs the default project profile in `reference-audit.md#audit-project-profile`). > **Refute panel:** shared - 3 skeptics per candidate, survive on >= 2 of 3 > (`reference-audit.md#audit-refute`). This pack does not opt out. Use each row as the `{{lens}}` / `{{lens_question}}` of `audit-finder.md`, one fin...Votes: 0GitHub stars: 23
- Bugs> **Status:** Fixed > **Verification depth:** functional (resolved from an EMPTY fixture root, so only the loaded skill dir can satisfy it; absent-path control and the derivation both pinned) > **Severity:** High > **Points:** 3 > **Affects:** .claude/skills/sdlc-studio/scripts/lib/sdlc_md.py, .claude/skills/sdlc-studio/scripts/readiness.py, .claude/skills/sdlc-studio/scripts/tests/test_detector_owed.py, .claude/skills/sdlc-studio/scripts/tests/test_sdlc_md.py, .claude/skills/sdlc-studio/scri...Votes: 0GitHub stars: 23
- Epics> **Status:** Done > **Derived Point Total:** 1 > **Parent:** CR0370 > **Created:** 2026-07-20 > **Created-by:** sdlc-studio new > **Raised-by:** sdlc-studio; agent; v1 > **Size:** SVotes: 0GitHub stars: 23
- Bugs> **Status:** Fixed > **Groomed:** 2026-10-01 (D0291) - premise reproduced at HEAD: `grep -rn CLAUDE_CONFIG_DIR tools/ evals/` finds nothing; `eval_run.py setup` prints `--- WORKER PROMPT (fresh session, skill installed) ---` and evals/README.md step 2 says only `with the candidate skill installed`. Narrowed: setup always prints the isolated command, so the personal-copy diff warning is dropped (one more check, no extra value), and setup copies no credentials > **Severity:** Medium > **Points...Votes: 0GitHub stars: 23
- Bugs> **Status:** Open > **Severity:** Medium > **Points:** 3 > **Affects:** .claude/skills/sdlc-studio/scripts/verify_ac.py, .claude/skills/sdlc-studio/scripts/tests/test_verify_ac.py, .claude/skills/sdlc-studio/reference-verify.md, changelog.d/BG0961.md > **Evidence:** Field report, 2026-10-07, grooming Sprint 0 stories on sdlc-studio-lens. validate.py instructions forbids a tool-specific skill path in committed instructions (BG0853). A story Verify line that has to run a skill script has nowhe...Votes: 0GitHub stars: 23