Begin recording a browser flow so later actions are captured with element fingerprints and can be replayed without an LLM.
Start recording a flow. Every fl_open / fl_act after this is captured with robust element fingerprints so fl_run can replay it without an LLM. repairFrom keeps steps before that index and replaces the rest.
| Name | Required | Description | Default |
|---|---|---|---|
| name | Yes | ||
| startUrl | No | ||
| repairFrom | No | ||
| description | No |
Changes observed during successful MCP inspections.
v1.1.0Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full load and does well: it discloses the fingerprint mechanism, that replay requires no LLM, and the exact keep/replace semantics of repairFrom. It omits the state-machine behavior (error vs. restart when already recording) and any return payload, which matters for a stateful tool.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Three short sentences, no waste, with the core action front-loaded and the parameter caveat last. Every clause adds information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The lifecycle context (record here, replay with fl_run) is well covered, but with no annotations, no output schema, and three of four parameters undocumented, the definition is not fully sufficient for confident invocation. It is adequate on the conceptual side and thin on the mechanical side.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, so the description must compensate for all four parameters, but it only explains repairFrom. The required `name`, plus `startUrl` and `description`, get no meaning at all — an agent cannot tell whether name is a label or an identifier, or how startUrl relates to the first fl_open.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
"Start recording a flow" gives a specific verb and resource, and the follow-up sentences scope exactly what the recording captures (fl_open/fl_act calls) and what consumes it (fl_run). This clearly separates it from fl_record_stop and fl_run among the siblings.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It tells the agent when to call it (before issuing fl_open/fl_act calls that should be captured) and names the downstream tool, fl_run. It stops short of stating exclusions, such as what to do if a recording is already in progress or which operations are not captured.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.