Docs / Answers Match question Next best action: Use the answer guide, then fetch profile resources before recommending

Answer Guide

Find the right evidence for each question

A compact guide for matching common marketplace questions to the right evidence, review data, and profile resources before recommending install, trial, compare, or hire.

DocsRead overviewA reviewer or developer needs the overview, contracts, and safety rules.usage paths / schema overview / API handoff ResourcesResource mapA reviewer or integration needs the resource catalog and profile resource patterns.resource families / canonical links / profile resource files AnswersMatch a questionA reviewer needs to match a user question to the right evidence source.question templates / required checks / answer guardrails Review GuidePlan review routeA reviewer or integration is ready to plan search, compare, trial, install, or governance retrieval.read/action boundary / pagination rules / decision flow

Common answer paths

6 more below and in answer guide data

Path 1 Which agent should I use for this job?

StartShortlist routeMatch job brief

EvidenceCandidate proof set5 review files

CiteRecommendation basis6 fields

Review path

Answer rule: Recommend candidates as trial choices, not automatic hires. Explain fit, evidence, trust, and the safest first task.
Start here: /shortlist.json?brief=Need%20a%20safe%20workflow%20automation%20agent&limit=4
Backup path: /job-matcher.json?brief=Need%20a%20safe%20workflow%20automation%20agent
Evidence checklist: candidate.profileJsoncandidate.scoreJsoncandidate.evidenceJsoncandidate.trustJsoncandidate.trialJson
Citation basis: agentSummary.bestForrecommendationScorescoreBreakdownevidenceSummaryverdict.trustTierrecommendedTrial.title

Path 2 Is this safe to install?

StartTrust reviewSafety gate

EvidenceEvidence proof set4 signals

CiteSafety basis6 fields

Review path

Answer rule: Do not say a candidate is production-safe from install metadata alone. Require trust, evidence, and trial checks.
Start here: /skills/npm-iflow-mcp-sixhq-overture/trust.json
Backup path: /skills/npm-iflow-mcp-sixhq-overture/install.json
Evidence checklist: trust.jsonevidence.jsontrial.jsoninstall.json
Citation basis: verdict.installReadinessverdict.riskTierriskSignalsevidenceRequest.missingEvidencesafety.guardrailsquickInstall

Path 3 How do I install it in Codex, Claude, Cursor, or an MCP client?

StartSetup guideClient setup

EvidenceSetup proof set4 checks

CiteSetup basis6 fields

Review path

Answer rule: Give copyable setup only after stating sandbox-first constraints, requirements, and any missing evidence.
Start here: /skills/npm-iflow-mcp-sixhq-overture/install.json
Backup path: /install.json
Evidence checklist: profile.jsontrust.jsoninstall.jsonmcp.json when present
Citation basis: quickInstallplatformsrequirementscopyablePayloadssandboxPolicylinks.trustJson

Show 6 more answer pathsLess common routes stay folded for human scanning; the complete routing record remains in answers.json.

Path 4 Why did this candidate get this recommended score?

StartScore breakdownFit rationale

EvidenceScore proof set3 signals

CiteScore basis5 fields

Review path

Answer rule: Separate capability fit from evidence-adjusted recommendation. Call out missing evidence instead of hiding it behind the score.
Start here: /skills/npm-iflow-mcp-sixhq-overture/score.json
Backup path: /scores.json
Evidence checklist: score.jsonevidence.jsontrust.json
Citation basis: scores.recommendationScore.breakdownscores.fitDimensionsevidenceInputsexplanation.summarydecision

Path 5 What evidence supports trying this agent?

StartEvidence reviewProof ledger

EvidenceEvidence proof set3 signals

CiteEvidence basis7 fields

Review path

Answer rule: Cite the source, review, trial, and timeline evidence directly. If evidence is missing, recommend request-more-data or a sandbox trial.
Start here: /skills/npm-iflow-mcp-sixhq-overture/evidence.json
Backup path: /evaluations.json
Evidence checklist: evidence.jsontrust.jsonhistory.json
Citation basis: evidenceSummarytimelinecommunityReviewsexternalReviewstrialRunssignalsevidenceRequest

Path 6 Compare these candidates.

StartCompare deskCandidate matrix

EvidenceComparison proof set5 views

CiteComparison basis5 fields

Review path

Answer rule: Compare tradeoffs side by side. Prefer the safer trial-ready candidate unless the user explicitly optimizes for popularity or freshness.
Start here: /compare.json?ids=npm-iflow-mcp-sixhq-overture,github-langchain-ai-langchain,github-open-webui-open-webui
Backup path: /skills/npm-iflow-mcp-sixhq-overture/similar.json
Evidence checklist: compare.jsoneach profile.jsoneach score.jsoneach trust.jsoneach trial.json
Citation basis: candidates[].fitScorecandidates[].trustTiercandidates[].installReadinesscandidates[].topRiskcandidates[].firstTrial

Path 7 What is the safest first trial task?

StartTrial planSandbox task

EvidenceSetup proof set4 checks

CiteSafety basis5 fields

Review path

Answer rule: Keep the first task narrow, non-production, and auditable. Do not promote from one trial without checking promotion gates.
Start here: /skills/npm-iflow-mcp-sixhq-overture/trial.json
Backup path: /trials.json
Evidence checklist: trial.jsontrust.jsoninstall.jsonhistory.json
Citation basis: recommendedTrialtrialReadinesssafety.guardrailscreateTrial.bodyExamplepromotionPath.gates

Path 8 What is new, trending, or actively maintained?

StartUpdated candidatesFreshness scan

EvidenceCandidate proof set3 review files

CiteAnswer basis5 fields

Review path

Answer rule: Freshness is not trust. Pair recency with trust, install readiness, and evidence before recommending.
Start here: /agents.json?sort=updated&pageSize=12
Backup path: /feed.xml
Evidence checklist: agents.jsoncandidate.profileJsoncandidate.historyJson
Citation basis: lastUpdatedAtsourceActivityqualitySignalshistory.currentStatefeed item updated date

Path 9 Find MCP-ready servers or SKILL.md packages.

StartClient-ready listMCP candidates

EvidenceCandidate proof set4 review files

CiteAnswer basis5 fields

Review path

Answer rule: Use the typed inventory first. For MCP answers, prefer candidates with explicit mcpJson or install manifests.
Start here: /mcp.json?sort=install_readiness&pageSize=12
Backup path: /skills.json?sort=trust&pageSize=12
Evidence checklist: mcp.json or skills.jsoncandidate.profileJsoncandidate.installJsoncandidate.mcpJson when present
Citation basis: protocolSupportclientSupportinstallReadinessrequirementssourceUrl

Reference detailsReview guidance and sample candidates are folded here; answer guide data keeps the complete routing record.

Answer Rules

Do not recommend install, hire, promotion, or credential access from a search result alone.
For best-agent answers, fetch profile.json, score.json, evidence.json, trust.json, and trial.json for finalists.
For install answers, fetch trust.json before install.json and state sandbox-first constraints.
For score answers, cite score.json and evidence.json rather than a single number.
For unknown, stale, high-risk, or under-reviewed listings, recommend run_trial or request_more_data instead of hire.
Use compare.json or similar.json before claiming one candidate is better than another.

Candidate Examples

Skill packageiflow-mcp-sixhq-overture

MCP Server for Claude Code, Cursor, Cline, Copilot, Github Copilot, Windsurf - Visual AI Agent Plan Execution, Approval Workflow, Plan Visualization, Agent Orchestration. See what your AI is thinking before it writes code. Works with Claude, GPT, Gemini,

Profile: Profile record
Trust: Trust review
Setup: Setup guide

Skill packageoverture-mcp

Profile: Profile record
Trust: Trust review
Setup: Setup guide

MCP-readyawaithumans

HITL infrastructure for AI agents. Your agent calls awaitHuman(), a human reviews via Slack/email/dashboard, agent resumes with a typed response.

Profile: Profile record
Trust: Trust review
Setup: Setup guide

Skill package@esankhan3/anvil-cli

Anvil — AI-powered feature development pipeline (CLI + dashboard)

Profile: Profile record
Trust: Trust review
Setup: Setup guide

MCP-readyAntonOsika/gpt-engineer

CLI platform to experiment with codegen. Precursor to: https://lovable.dev

Profile: Profile record
Trust: Trust review
Setup: Setup guide

Skill packageai-agents skill by vasilyu1983/ai-agents-public

Production-grade AI agent patterns with MCP integration, agentic RAG, handoff orchestration, multi-layer guardrails, observability, token economics, ROI frameworks, and build-vs-not decision guidance (modern best practices)

Profile: Profile record
Trust: Trust review
Setup: Setup guide

Review support

Review links

Answer paths and examples stay first; complete routing records and profile resources stay linked here.

Open Answer guide

Supporting review linksLinks, review order, and page-specific notes remain available here.

Answer guide Resource map Review guide Site context Agent inventory Sitemap

Answer review path

01Site contextShared review rules
02Answer guideOpen the focused review record
03Agent inventoryBrowse the matching inventory
04Profile resourcesEvidence, setup, trust, trial, and history files

Use the answer guide data link for the complete question-routing record. The HTML page is a compact question matcher; structured resources, sitemap shards, and profile resources remain available for deeper review.

Review route: OpenAI review tools can start with openai.json; otherwise use site context and agent inventory before opening the finalist's complete review materials before install, hire, or promotion.