Skip to main content
Glama

Jailbreak Detection Check

dcl_evaluate_jailbreak

Detect jailbreak attempts in agent responses by checking for forbidden phrases like 'ignore previous instructions' and 'jailbreak'. Returns COMMIT if safe, or NO_COMMIT with reason if a match is found.

Instructions

PRE-ACTION Instruction Adherence Check ($0.02). Runs the "anti_jailbreak" policy: a substring check against 6 forbidden phrases ("ignore previous instructions", "jailbreak", "bypass safety", "pretend you are", "act as if", "DAN") with a 0.8 minimum-confidence threshold — each match costs 0.4 confidence. Returns COMMIT if no phrase matches and confidence stays at or above 0.8, otherwise NO_COMMIT with reason listing the matched phrase(s). Use this as a targeted, cheaper check when the concern is specifically prompt-injection / persona-hijack risk; use dcl_evaluate_strict instead when you also want safety- and default-policy phrases covered in the same call.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
agent_idYesIdentifier of the agent that produced the response.
responseYesThe agent or LLM response text to check for jailbreak attempts.
payment_idNoOptional payment identifier returned by a previous call when payment is required

Output Schema

TableJSON Schema
NameRequiredDescriptionDefault
reasonYesHuman-readable explanation of why the verdict was reached.
tx_hashYesHash of this record in the tamper-evident audit chain.
verdictYesCOMMIT if the response passed policy checks, otherwise NO_COMMIT.
seal_textYesHuman-readable Leibniz Layer verification seal.
timestampYesUnix timestamp when this record was sealed.
confidenceYesConfidence score of the verdict, from 0.0 to 1.0.
drift_modeYesCurrent drift status: NORMAL, WARNING, ESCALATION, or BLOCK.
input_hashYesHash of the evaluated response (raw content is never stored).
verify_urlYesPublic URL to independently verify this seal.
chain_indexYesSequential index of this record in the audit chain.
drift_scoreYesZ-score measuring deviation of the recent commit rate from baseline.
policy_versionYesVersion of the policy that was applied.
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

The description reveals behavioral details beyond annotations: the cost ($0.02), the exact algorithm (substring check, 0.8 confidence threshold, 0.4 confidence penalty per match), and the return format (COMMIT/NO_COMMIT with reason listing matched phrases). This gives substantial context that annotations do not provide.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is compact and well-structured, packing cost, algorithm, return behavior, and usage guidance into two sentences with no unnecessary words. It is dense but highly readable.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the output schema exists, the description does not need to explain return values in detail. It covers all essential aspects: purpose, cost, algorithm, when to use, and how it differs from alternatives. No gaps remain.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema already has 100% description coverage for all three parameters, so the baseline is 3. The description does not add parameter-specific semantics beyond what the schema provides; it focuses on the algorithm and use case rather than on the parameters themselves.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool runs an 'anti_jailbreak' policy with substring checks against 6 forbidden phrases and returns COMMIT or NO_COMMIT. It distinguishes itself from dcl_evaluate_strict, specifying the exact scope (prompt-injection/persona-hijack) and differentiating from sibling tools.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It explicitly provides usage guidance: use this targeted, cheaper check for prompt-injection/persona-hijack risk, and use dcl_evaluate_strict when broader safety/default-policy coverage is needed. This clearly states when to use this tool versus the alternative.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/Fronesis-Labs/dcl-webhook'

If you have feedback or need assistance with the MCP directory API, please join our Discord server