hafizmirhamza276-lab

backend-agentic-marketplace

Explore a backend codebase ONCE with breadth+depth sub-agents, then write a durable .claude/explorer/ memory so any future session understands the code (what / why / how) without re-exploring.

Spec-driven agentic implementation for Claude Code. Reads explorer codebase memory as ground truth, plans a requested change with clarity + technical-plan gates (9/10 required), implements only what the spec says, runs hybrid QA, and keeps the durable memory in sync. Includes a reproduce-first BUG-FIX MODE (failing reproduction + characterization tests + deterministic regression gate) for vague, repro-less bug reports.

Conductor plugin: sequences the explorer -> builder flow (running exploration first when memory is missing/stale, then the spec-driven build or BUG-FIX MODE) and gates the result behind a deterministic RELEASE GATE — fresh explorer memory, builder done with a full task coverage map, bug-fix verification net green, CHANGELOG present, and (when present) auditor 0-high. Advisory by default; enforce via settings.enforce_release or PIPELINE_ENFORCE=1. The gate is pure shell/awk, so it behaves identically on python-less / Windows hosts.

Audits this marketplace against its own F1–F13 failure classes: deterministic pure-shell/awk regression detectors (fail-open, path traversal, broken hook contracts, manifest + vendored-lib drift, notebook/stdin/timeout/line-ending gaps) plus breadth+depth review sub-agents. Writes .claude/auditor/FINDINGS.md and records a HIGH count in the STATUS contract so the pipeline release gate enforces 0-high. Silent on a clean tree; no python dependency.

Reviews a CHANGE (the working-tree diff vs HEAD) against the explorer MEMORY.md invariants and risk map, the approved PLAN scope, and the codebase's surviving callers. Deterministic pure-shell/awk checks — R1 caller-integrity & R2 convention-regression (BLOCKING), R3 risk-touch & R4 scope-discipline (CONCERN) — plus breadth+depth review sub-agents. Writes .claude/reviewer/REVIEW.md and records a BLOCKING count in the STATUS contract so the pipeline release gate enforces 0-blocking. Orthogonal to the auditor (change-vs-invariants, not whole-tree regression). Silent on a clean/in-spec change; no python dependency.

Assesses DEPLOY/RELEASE readiness of the codebase. Deterministic pure-shell/awk checks kept deliberately lean — O1 test-ledger (a recorded RED build/test → BLOCKING; no ledger → CONCERN) and O2 version-consistency (a marketplace entry's version ≠ that plugin's plugin.json version → CONCERN) — plus breadth+depth ops sub-agents that inventory the CI/deploy/observability surface and reason about readiness as advisory CONCERN/NOTE findings. Writes .claude/ops/OPS.md and records a BLOCKING count in the STATUS contract so the pipeline release gate enforces 0-blocking. Fuzzy deploy judgment lives in the agents, never a false-firing detector. Silent on a verified, consistent tree; no python dependency.

Always-on "write the least code that fully works" discipline. A SessionStart hook injects the full decision LADDER (skills/minimal-code) and a UserPromptSubmit hook re-injects a compact reminder every turn, so the agent ships the SMALLEST code that satisfies the requirement WITHOUT cutting validation, error handling, security, or accessibility. Both hooks are node-guarded and fail-quiet (node is OPTIONAL — absent means the hooks stay silent and the skill still works on demand) and INJECT CONTEXT ONLY (no gate decision; every gate in this marketplace stays pure shell/awk). Intensity off|lite|full|ultra via /minimize, mirrored into STATUS.json. YAGNI applies only to UNREQUESTED scope — anything the builder PLAN or the user mandates stays in-scope. Ladder adapted from Ponytail by Dietrich Gebert (MIT).