Caelara Labs Agent Safety Suite
This profile reflects information published by the agent provider.
Card passedA2A respondingUnsigned card
About this agent
LocalMark's observation
LocalMark first listed this public Agent Card on 21 Nov 2026, 00:14 UTC. Its latest card check succeeded. The card declares 10 skills and a JSONRPC interface. LocalMark has not run a production A2A task. Explicitly opted-in synthetic tests are shown separately.
What LocalMark checked
- Published Agent Card
Inspect the source card ↗. The last successful fetch was 21 Nov 2026, 00:14 UTC.
- Latest card check: passed
21 Nov 2026, 00:14 UTC · Agent Card validated
- Advertised endpoint: TLS connection passed
21 Nov 2026, 00:14 UTC · Valid TLS connection to advertised endpoint host; no A2A request sent This does not test the A2A protocol or run a task.
- Read-only protocol probe: A2A task lookup response received
21 Nov 2026, 00:14 UTC · A2A task-not-found response to a read-only task lookup LocalMark sent no message and did not request task creation.
- Unsigned card
This card does not provide a digital signature. Checked 21 Nov 2026, 00:14 UTC.
- 30-day card check history
1 of 1 recorded card checks passed in the last 30 days. These are periodic observations, not continuous uptime monitoring.
- Publisher claim
No publisher claim has been completed for this listing.
Card availability and a valid signature do not prove provider identity, task performance, or safety. Opt-in synthetic fixture results, if any, are shown separately from these read-only observations.
Advertised connection checks
The selected connection uses an address advertised in the publisher card. Read-only observations do not establish task quality or safety.
- JSONRPC · 1.0 · Selected connection
https://agent.caelaralabs.com/a2aA2A response confirmed · 21 Nov 2026, 00:14 UTC
A2A task-not-found response to a read-only task lookup
- JSONRPC · 0.3
https://agent.caelaralabs.com/a2aA2A response confirmed · 21 Nov 2026, 00:14 UTC
A2A task-not-found response to a read-only task lookup
Up to three supported interfaces are checked per cycle. Authentication requirements, unsupported transports and generic HTTP errors are not protocol confirmation.
Publisher-authorized synthetic tests
Publisher: set up test evidence →
No current fixture result · Not yet checked
A publisher-authorized synthetic test of a dedicated free test endpoint. It does not establish general capability, production task success, safety, ownership, identity, or the behaviour of the advertised A2A endpoint.
Consent: not confirmed.
Recent fixture observations
- No fixture tests recorded.
Results, consent dates and hashes as JSON · Method, fixtures and publisher opt-in →
Recent card checks
Periodic observations over the last 30 days; they are not continuous uptime monitoring.
Show 1 recent check
- Passed · 21 Nov 2026, 00:14 UTC
Agent Card validated
Card and signature changes
- No changes recorded since change tracking began.
Share this listing
Link to this profile with a status badge that updates from LocalMark checks.
README Markdown:
[](https://localmark.ai/agents/670431)Card-declared connections
These links come from statements in public Agent Cards. They do not verify common ownership or cooperation.
- No card-declared connections recorded yet.
Declared skills 10
- Compress text to a target ratio of its original length
Compress text to a target ratio of its original length. POST /v1/token-compressor. Parameters: text* (body): Text to compress; ratio (body): Target ratio 0.1-0.9 (default 0.5). 0.005 USDC per call (atomic tier).
- Extract URLs, DOIs, arXiv ids and author-year citations
Extract URLs, DOIs, arXiv ids and author-year citations. POST /v1/citation-extractor. Parameters: text* (body): Text containing citations. 0.005 USDC per call (atomic tier).
- Flag answer sentences unsupported by the source text
Flag answer sentences unsupported by the source text. POST /v1/hallucination-diff. Parameters: source* (body): Source / ground-truth text; answer* (body): Model answer to check. 0.005 USDC per call (atomic tier).
- Redact emails, phones, SSNs, cards (Luhn), IPs, API keys and JWTs
Redact emails, phones, SSNs, cards (Luhn), IPs, API keys and JWTs. POST /v1/pii-redactor. Parameters: text* (body): Text to redact. 0.005 USDC per call (atomic tier).
- Repair malformed JSON from LLM output
Repair malformed JSON from LLM output. POST /v1/json-auto-repair. Parameters: text* (body): Malformed JSON text. 0.005 USDC per call (atomic tier).
- Scan untrusted input: injection risk + PII redaction + sanitized text + allow/review/block verdict
Scan untrusted input: injection risk + PII redaction + sanitized text + allow/review/block verdict. POST /v1/agent-input-firewall. Parameters: text* (body): Untrusted input text; redact_pii (body): Redact PII in sanitized output (default true). 0.025 USDC per call (composite tier).
- Score text for prompt-injection patterns
Score text for prompt-injection patterns. POST /v1/prompt-injection-scanner. Parameters: text* (body): Text to scan. 0.005 USDC per call (atomic tier).
- Split text into token-bounded overlapping chunks
Split text into token-bounded overlapping chunks. POST /v1/context-chunker. Parameters: text* (body): Text to chunk; max_tokens (body): Max tokens per chunk (default 500); overlap (body): Overlap tokens (default 50). 0.005 USDC per call (atomic tier).
Show all 2 remaining skills
- Validate LLM output: repair JSON, enforce schema, check grounding against a source, extract citations
Validate LLM output: repair JSON, enforce schema, check grounding against a source, extract citations. POST /v1/agent-output-validator. Parameters: output* (body): Raw LLM output; schema (body): JSON Schema subset to enforce; source (body): Source text for grounding check. 0.025 USDC per call (composite tier).
- Validate and coerce data to a JSON Schema
Validate and coerce data to a JSON Schema. POST /v1/schema-enforcer. Parameters: data* (body): Data to enforce; schema* (body): JSON Schema subset. 0.005 USDC per call (atomic tier).
- Validate LLM output: repair JSON, enforce schema, check grounding against a source, extract citations