vuongdam2k01
UserA Claude Code plugin that drives a raw SaaS idea to a locked MVP spec — competitive scan, market validation, feasibility checks, positioning — and refuses to call anything "validated" that isn't.
Categories
Indexed Skills (40)
amend-blueprint
Record a mid-build discovery against a locked implementation blueprint - run the founder-answered scope test, then either write an immutable Blueprint Amendment Record (in-scope fix) or route to declare-drift (scope-level change). The locked blueprint is never edited; current truth = locked blueprint + amendment log.
declare-drift
Record that a post-LOCK SaaS idea's product reality has drifted from its recorded truth - appends to the drift inbox and arms the reconcile boundary. Use when the user mentions having shipped, changed, removed, or repriced something since LOCK, without wanting a full reconcile right now. Also handles a founder's controlled MVP release declaration (--release) when V1 was deferred.
gate-check
Run the three-layer gate check for a SaaS validation idea - formal contract checks, adversarial gatekeeper review, user approval. Use when a stage's artifacts are complete and a gate verdict is needed, or when the user asks to check a gate.
handoff-to-build
Ship a locked pack + blueprint to a build repository that is SEPARATE from the idea workspace, so the coding agents there arrive already knowing the specs exist, that they are frozen, how to resolve an id, and where a spec defect goes. Only for the two-repo case - when the code lives in the same repo as the idea workspace there is nothing to generate, because the plugin already does this natively. Use after gate BP passes, after every blueprint amendment, and whenever a build session in that repo reports it had to guess.
method-rules-artifact-schema
Canonical artifact frontmatter and evidence-ledger row format for the SaaS validation pipeline. Load before writing any artifact under ideas/<slug>/.
method-rules-gate-contracts-bp
Normative contract for gate BP (stage 6, the implementation blueprint) - coverage predicates, interaction and subsystem layers, surfaces, compliance, amendment boundary. Load only when working on stage 6 or checking gate BP.
method-rules-gate-contracts
Normative per-gate contracts for the SaaS validation pipeline - required artifacts, prerequisites, metrics, OPEN rules, and the exact pack-verdict predicate. Load before any gate work.
method-rules-maintenance-rules
Post-LOCK maintenance rules for the SaaS validation pipeline - cycles, drift declaration, reconciliation, validation runs, health criteria. Load for any post-LOCK work.
method-rules-state-schema
Canonical state.json schema (v1.2) for the SaaS validation pipeline - exact shape, field rules, statuses, migration, cycles. Load before creating or updating any ideas/<slug>/state.json.
method-rules
Core operating rules for the SaaS idea validation pipeline - evidence grading (A/B/C/D), GUESS labeling, traceability, pre-registered thresholds, gate discipline. Load whenever working on any artifact under ideas/<slug>/ or evaluating evidence for this pipeline.
new-idea
Start validating a new SaaS idea - creates the idea workspace and begins stage 0 framing.
reconcile
On-demand reconciliation for a post-LOCK SaaS idea - compare declared/observed product reality against the idea's direction artifacts, re-baseline current truth, and route unvalidated drift to validation runs. Use when the user returns after product changes, declares drift, or asks to sync the idea's docs with reality.
run-validation
Execute and adjudicate a signed post-LOCK validation run for a SaaS idea - gather evidence inside the confirmation window, compute the verdict against the pre-signed spec, write the immutable run report, and update claim statuses. Use when a signed validation-run spec's window is open or has closed and needs a verdict.
setup-audit
Detect which optional integrations are actually callable and record the capability profile with execution rungs. Use any time, and whenever a stage needs to pick between enhanced-auto, baseline-auto, or handoff execution. Integrations are never required.
spec-gap
Use when the locked spec under ideas/<slug>/ does not answer an in-scope question, contradicts itself, or turns out to be wrong once real code exists - an unhandled case, a provider behaving differently than documented, two specs that cannot both be satisfied, a stated limit that is impossible. Triages product versus technical, checks the places answers usually hide, and either resolves it or drafts the amendment. Reach for it BEFORE deciding an unspecified product behaviour during a build.
spec
Resolve a locked product-spec id - fs-NN, AC-NN-n, ST-<entity>-n, INV-n, JOB-n, CAP-NN-n, EV-n, DR-n, DOD-n, MSP-n, SC-n, DF-n, BA-nnn - to the exact file, section and text that DEFINES it. Use whenever such an id appears in a spec, a task, a test name, a comment or a commit message, and before implementing any behaviour that references one. The specification under ideas/<slug>/ is frozen and authoritative, so the answer is looked up, never inferred.
stage-0-framing-templates
Artifact templates for stage 0 of the SaaS validation pipeline. Load when writing that stage's artifacts.
stage-0-framing
Stage 0 of the SaaS validation pipeline - turn a raw idea into a testable hypothesis set. Use when an idea in ideas/<slug>/ is at stage 0 - problem framing, lean canvas, market type, beachhead ICP, assumption map, kill criteria.
stage-1-competitive-templates
Artifact templates for stage 1 of the SaaS validation pipeline. Load when writing that stage's artifacts.
stage-1-competitive
Stage 1 of the SaaS validation pipeline - competitive and alternatives scan. Use when an idea in ideas/<slug>/ needs the 5-tier competitor map, competitor profiles, negative-review mining, and the market verdict. Runs right after task 0.1, before any interviews/mining for stage 2.
stage-2-validate-templates
Artifact templates for stage 2 of the SaaS validation pipeline. Load when writing that stage's artifacts.
stage-2-validate
Stage 2 of the SaaS validation pipeline - market validation through three sequential gates - V1 problem evidence, V2 solution direction, V3 money commitment. Use when an idea in ideas/<slug>/ has passed gate F/C and needs problem mining or interviews, solution directions with the ChatGPT-gap test, mocks, landing/pre-sell kits, or willingness-to-pay analysis.
stage-3-verify-templates
Artifact templates for stage 3 of the SaaS validation pipeline. Load when writing that stage's artifacts.
stage-3-verify
Stage 3 of the SaaS validation pipeline - solution verification, run IN PARALLEL with stage 2. Use when an idea in ideas/<slug>/ has a deadly feasibility assumption (always true for AI-core products) - spike/PoC, error analysis, eval harness, unit economics (gate R1), and value-delivery verification (gate R2).
stage-4-positioning
Stage 4 of the SaaS validation pipeline - positioning per April Dunford's component order. Use when an idea in ideas/<slug>/ has V1 evidence (customer words) and needs its positioning thesis - competitive alternatives from evidence, unique attributes, value and proof, target segment, market category, copy test, pitch.
stage-5-scope-lock-templates
Artifact templates for stage 5 of the SaaS validation pipeline. Load when writing that stage's artifacts.
stage-5-scope-lock
Stage 5 of the SaaS validation pipeline - lock the MVP scope and produce the final MVP Pack. Use when an idea in ideas/<slug>/ has V3 and R2 outcomes (passed or accepted-open) and locked positioning - core loop with traceability, aha event, cut list, technical design contract, definition of done, cold-start test.
stage-6-blueprint-templates
Artifact templates for stage 6 of the SaaS validation pipeline. Load when writing that stage's artifacts.
stage-6-blueprint
Stage 6 of the SaaS validation pipeline - turn the locked MVP Pack into a complete implementation blueprint before any product code. Use when an idea in ideas/<slug>/ has passed LOCK (or recorded an unvalidated-build decision) and the founder is preparing to build - feature specs, field-level data schema, UX spec, API and integration contracts, NFRs, test plan, build plan, level-2 cold-start test, gate BP.
status
Show validation pipeline status for ideas in this workspace - stage, gates, waiting-on items, kill-criteria deadlines.
switch-mode
Switch a SaaS validation idea between analysis mode and market-evidence mode, with precondition checks and journaling.
object-object
Resolve a frozen product-spec id (fs-NN, AC-NN-n, INV-n, JOB-n, ST-<entity>-n, CAP-NN-n, EV-n, DR-n, DOD-n, MSP-n, DF-n, E-nnn) to the exact file, section and text that defines it. Use whenever such an id appears in a spec, task, test name, comment or commit message, and before implementing any behaviour that references one — the specification in {{SPEC_ROOT}}/ is frozen and authoritative, so the answer is looked up, never inferred.
sag-setup
First-run setup for this plugin on this machine — verify the SAG instance, mint tokens, wire the agent tool, and either bootstrap a new scope or join one that already exists. Runs only when the user types /sag-setup or explicitly asks to set the plugin up.
stage-4-positioning-templates
Artifact templates for stage 4 of the SaaS validation pipeline. Load when writing that stage's artifacts.
sag-source-admin
Create, update, or delete a SAG source (the top-level container documents belong to). Destructive and rarely needed — only invoked explicitly by a human, never by model judgement.
sag-sync-project
Batch-sync every file matching the SAG manifest's include patterns to the knowledge base. Only invoked explicitly by a human or the orchestrator profile — never triggered automatically by the model on its own judgement.
sag-knowledge
Use when searching, browsing, citing, or reading documents from the SAG knowledge base through the read-only MCP server "sag" — before architecture decisions, when starting a new task in a domain that may already have documented decisions, or when a claim needs a cited source.
sag-maintain
Use when asked to check the health of the SAG knowledge base, review documents stuck in failed status, find orphaned or duplicate documents, or run the periodic self-gate review of auto-published documents. Typically invoked on a schedule (orchestrator profile / scheduled task) rather than ad hoc.
sag-publish
Use immediately after creating or editing a document (markdown, or anything else — requirement, ADR, design doc, research report, runbook, postmortem, API contract, PDF/DOCX you analyzed) — self-assess whether the content is durable, shared knowledge worth publishing to the SAG knowledge base, and call sag_publish or sag_publish_content if so. Also use when the user explicitly asks to publish a document to the knowledge base.
sag-status
Use when asked whether the SAG plugin is set up correctly on this machine, which scope it is pointed at, whether it shares state with other agent hosts, or whether committed files are missing from the knowledge base. Read-only diagnostics — changes nothing.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.