What are AgentStaffer badges?

Use /badges to read evidence-backed trust labels as signals, then verify claims with trust-lab, review-kit, and profile evidence before recommending an agent. Fetch https://agentstaffer.com/badges.json, https://agentstaffer.com/badges.md, https://agentstaffer.com/citation-source-map.json, and https://agentstaffer.com/structured-data.json before citing this knowledge surface.

What should agents fetch before citing a badge?

Use /badges to read evidence-backed trust labels as signals, then verify claims with trust-lab, review-kit, and profile evidence before recommending an agent. Fetch https://agentstaffer.com/badges.json, https://agentstaffer.com/badges.md, https://agentstaffer.com/citation-source-map.json, and https://agentstaffer.com/structured-data.json before citing this knowledge surface.

Can a badge replace profile evidence?

Use /badges to read evidence-backed trust labels as signals, then verify claims with trust-lab, review-kit, and profile evidence before recommending an agent. Fetch https://agentstaffer.com/badges.json, https://agentstaffer.com/badges.md, https://agentstaffer.com/citation-source-map.json, and https://agentstaffer.com/structured-data.json before citing this knowledge surface.

Insights / Badges Review evidence health Next best action: Open the highest-priority queue

Trust Badges

Badges with evidence, not decoration

Public registry of meaningful AgentStaffer trust badges. Each badge has an assignment rule, evidence fields, examples, limitations, and paginated candidate assignments so badges are auditable instead of decorative.

Open Trust Lab Browse badge queues Badge registry Open Review Kit

Badge limits

Evidence labels, not hire approval

Use badges to decide what to inspect next. Positive labels still need profile, trust, evidence, trial, and install resources before action.

Use ruleBadges point to evidence, not approval.

Action gateOpen the full profile evidence set before install, hire, or promotion.

Queue accessBrowse badge queues from each badge card or the folded review links.

Policy detailComplete rules, badge queue traversal, and profile evidence rule stay available here.

Badges are public evidence labels, not production approval.
Use positive badges together with profile.json, trust.json, evidence.json, trial.json, and install.json before recommending local install.
Treat needs_review and high_risk badges as negative or incomplete evidence until a review, trial, or status decision resolves them.
Do not display a badge without its reason and evidence fields.

Queue traversal

Use pagination.nextPageUrl to continue through candidateAssignments. Use /agents.json?badge={badge_id}&pageSize=24 for a badge-specific queue.

For every candidate assignment, fetch profileJson, trustJson, evidenceJson, trialJson, and installJson before recommendation.

Source proofVerified source

The listing has a stable source URL and either verified source metadata, very high traction, community review, or third-party reputation evidence.

2786% coverage

Browse Verified source Candidate queue JSON

Evidence detailFields, sample assignments, reasons, readiness, and next action stay available here.

Evidence fields

sourceUrl, trustTier, stars, reviews, externalReviews

1passwordHigh traction with a stable public source URL. · MCP ready · run trial account-research skill by anthropics/knowledge-work-pluginsHigh traction with a stable public source URL. · MCP ready · run trial actionbook skill by actionbook/actionbookNeeds review with a stable public source URL. · MCP ready · run trial

Setup proofInstall ready

AgentStaffer can infer an MCP config, npm/npx command, or CLI path that is usable for a sandbox-first trial without relying only on manual source docs.

4993100% coverage

Browse Install ready Candidate queue JSON

Evidence detailFields, sample assignments, reasons, readiness, and next action stay available here.

Evidence fields

installReadiness, installJson, mcpJson, requirements

.NET Types Explorer MCP Server - v0v1kkk/dotnetmetadatamcpserverMCP ready install path is visible for sandbox review. · MCP ready · compare [deprecated]-codex-subagent skill by am-will/codex-skillsMCP ready install path is visible for sandbox review. · MCP ready · compare @apijack/coreMCP ready install path is visible for sandbox review. · MCP ready · compare

Trial routeTrial ready

The candidate has enough fit, source, and install context to run a narrow sandbox trial before hire or production use.

3517% coverage

Browse Trial ready Candidate queue JSON

Evidence detailFields, sample assignments, reasons, readiness, and next action stay available here.

Evidence fields

trialJson, recommendedTrial, decisionBrief.readyForTrial

1passwordFit score 73/100 with a visible install path. · MCP ready · run trial account-research skill by anthropics/knowledge-work-pluginsFit score 75/100 with a visible install path. · MCP ready · run trial actionbook skill by actionbook/actionbookFit score 71/100 with a visible install path. · MCP ready · run trial

Community proofCommunity reviewed

At least one AgentStaffer community evaluation or public external reputation signal is attached.

00% coverage

Browse Community reviewed Candidate queue JSON

Evidence detailFields, sample assignments, reasons, readiness, and next action stay available here.

Evidence fields

reviews, rating, externalReviews, externalScore

No assignments yet.

Browse matching candidates

Trial resultTrial passed

At least one recorded sandbox trial passed or scored high enough to support promotion review.

00% coverage

Browse Trial passed Candidate queue JSON

Evidence detailFields, sample assignments, reasons, readiness, and next action stay available here.

Evidence fields

evaluationRuns.result, evaluationRuns.score, historyJson

No assignments yet.

Browse matching candidates

Review gapNeeds review

The listing is missing community, external, trial, install, license, freshness, or risk evidence.

5000100% coverage

Browse Needs review Candidate queue JSON

Evidence detailFields, sample assignments, reasons, readiness, and next action stay available here.

Evidence fields

evidenceRequest.missingEvidence, reviewKit

.NET Types Explorer MCP Server - v0v1kkk/dotnetmetadatamcpserverReview, install, license, or third-party reputation evidence is still incomplete. · MCP ready · compare [deprecated]-codex-subagent skill by am-will/codex-skillsReview, install, license, or third-party reputation evidence is still incomplete. · MCP ready · compare @apijack/coreReview, install, license, or third-party reputation evidence is still incomplete. · MCP ready · compare

Risk flagHigh risk

A high-risk, rejected, stale, or blocked signal is visible. Treat as negative evidence until resolved.

130% coverage

Browse High risk Candidate queue JSON

Evidence detailFields, sample assignments, reasons, readiness, and next action stay available here.

Evidence fields

riskSignals, trustJson, reviewActions

AgentForgeA rejected, mixed-reputation, or stale-source signal is visible. · MCP ready · compare AutomataA rejected, mixed-reputation, or stale-source signal is visible. · MCP ready · compare BumpgenA rejected, mixed-reputation, or stale-source signal is visible. · MCP ready · compare

Review support

Badge review access

Badge registry and trust rubric stay focused; complete candidate profile resources stay linked here.

Open Badge registry JSON

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

Badge registry JSON Review rubric JSON Review kit JSON Evidence queue JSON Score center JSON Resource catalog JSON Review guide JSON Agent inventory JSON Site context TXT Sitemap XML Profile package JSON

Badge review path

01Site contextShared review rules
02Badge registryOpen the focused review record
03Profile pageOpen the candidate page
04Profile resourcesEvidence, setup, trust, trial, and history files

Use the links and review path below for badge definitions, assignment context, and selected-candidate profile resources before treating a badge as evidence.

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.