Tickerr MCP

Quick Verdict

Tickerr MCP is a real installable MCP package backed by a live remote service. npm lists tickerr-mcp 1.1.2 as latest. It can act as a stdio wrapper around Tickerr’s API, while clients with remote MCP support can connect directly to https://tickerr.ai/mcp. Its scope includes AI tool status, incident history, model pricing, limits, performance, and agent-reported failures. Think of it as a status reference inside an assistant, not a model provider or an SLA-backed routing control plane.

Treat its output as provisional and informational. The live tools API returned 96 entries during review, while pages and package text use “90+.” The site and README advertise nine MCP tools, but the public source, README, and hosted descriptions are not perfectly aligned. npm metadata says MIT, yet the repository root has no retrievable LICENSE file. Incident reporting can include provider, model, error code, latency, region, and other operational fields, while retention, agent deduplication, consent ownership, and independent auditing remain unclear. Do not let it autonomously control production failover or procurement.

Best For

• AI application developers seeking a quick external clue during provider failures
• Cost researchers doing an initial pricing or free-tier scan before checking official documentation
• MCP evaluators testing how status and incident tools appear in an assistant conversation
• Operations and support teams using it as corroborating evidence alongside internal telemetry and official status pages
• Not ideal for organizations requiring an auditable SLA, documented data provenance, strict telemetry governance, or unattended production switching

Key Features

• Status lookup for monitored AI products, response time, check time, uptime, and active incidents
• Incident history with severity, duration, and affected components
• Pricing and limits across models, public plans, quotas, and free tiers
• Workload cost comparison from input and output token counts
• Performance references advertised by the hosted service, including latency and throughput, although package and hosted capabilities require contract testing
• Crowdsourced reports through report_incident, with aggregated signal and fallback suggestions
• Remote or stdio access: HTTP MCP directly, or npm 1.1.2 on Node.js 18+ as a wrapper that still calls Tickerr online

Use Cases

• Outage troubleshooting evidence checked against provider status pages and local metrics
• Pricing shortlists followed by verification of regional, cached, batch, and tiered prices in official documents
• Public limit questions where the account console remains the authority for actual quota
• MCP prototypes that test safe behavior when an external status service is unavailable
• Incident research over recent event history without treating the result as contractual availability proof

Pricing

Tickerr presents MCP reads and REST reporting as free and requiring no API key. npm identifies version 1.1.2 and Node.js 18+; the local package primarily forwards stdio MCP calls to hosted Tickerr endpoints. The public page does not establish an enterprise SLA, capacity commitment, durable export guarantee, or long-term compatibility policy. Confirm these in writing before creating a production dependency.

The license trail also needs clarification. npm package.json declares MIT, but the public GitHub root currently has no retrievable LICENSE text. A metadata field alone should not be used to infer licensing for every source file or historical release. Organizations modifying, redistributing, or bundling the package should wait for a license file or explicit permission.

Pros

• Straightforward remote MCP and npm stdio onboarding
• Status, incidents, pricing, limits, and cost comparisons in one conversational interface
• A live API genuinely returns a monitored tool directory, with 96 entries observed during review
• Read access without an API key is convenient for prototypes
• Crowdsourced reports can add a secondary clue beside official pages and internal monitoring

Cons

• npm 1.1.2, the website, README, and public source disagree on parts of the tool contract and status semantics
• The live count of 96 versus “90+” shows that coverage figures are approximate, not contractual
• npm says MIT but the repository lacks a LICENSE file
• Telemetry retention, consent, regional fields, identity deduplication, and withdrawal are insufficiently documented
• Data provenance, sampling, and independent validation are limited, so prices and performance need official confirmation
• Fallback recommendations omit business quality, residency, capacity, compliance, and contract constraints

Alternatives

Use OpenRouter or a controlled LiteLLM deployment when traffic actually needs routing. Feed those systems internal success rate, latency, quality, and budget policy. Tickerr fits better as a side-channel signal; n8n or Dify can add corroboration and approval so one external observation does not redirect production traffic.

FAQ

Is Tickerr MCP a real product?

Yes. npm hosts version 1.1.2, and tickerr.ai operates live MCP, status, and API endpoints. Being real does not make it mature; evaluate it as an early external information service.

Why does the API return 96 tools while documentation says 90+?

“90+” is approximate marketing language. On July 21, 2026, the live tools endpoint returned count: 96. Clients should read the current count and never encode the headline as a coverage guarantee.

Does it expose nine MCP tools?

The website and README say nine, but public source and hosted descriptions are not fully synchronized around get_model_performance, my_status, and status enums. List the tools actually visible to your client and contract-test critical calls.

Can Tickerr choose a production fallback automatically?

It should not be the sole decision maker. Its suggestion does not fully model application quality, context needs, residency, privacy, budget, quota, or contractual constraints. Use audited policy and multiple signals.

What does incident reporting transmit?

The public interface can submit provider, model, error code and type, latency, region, client tier, and recovery state. Tickerr says request content and personal data are excluded, but operational metadata can still reveal infrastructure and usage patterns.

Is the package MIT licensed?

npm metadata says MIT, but the repository currently lacks a root LICENSE file. Record the discrepancy for experiments and obtain auditable license text before organizational redistribution, modification, or supply-chain approval.

Bottom Line

Tickerr MCP 1.1.2 brings useful status, incident, pricing, and agent-report signals into assistants with minimal setup. However, 96 live entries versus “90+,” differences between the advertised nine tools and public source, the missing repository LICENSE, and unclear telemetry consent boundaries all make it a provisional information source. Use it for debugging, research, and prototypes, not direct production routing. Disable unapproved reporting, add timeouts and circuit breakers, timestamp every observation, and verify conclusions against official status pages and your own metrics.