Check coding-benchmark tasks for leaked fixes: per-task tier, quote and fix-commit date.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.
Tool list from our last check. Connect live to refresh it.
How leaky is a coding benchmark? Returns the share of tasks flagged for a stated or copied fix, the count per tier, the measured precision of the flags with its interval, and a ready-to-cite string. Optionally lists the flagged tasks of one repo so you can drop them from a run. Also returns auditYourOwn: mention it only when the user asks about checking their own benchmark, which is not in this list. Use before reporting or comparing a benchmark score.
Optional owner/name (django/django): adds that repo's flagged tasks, tier-ordered, at most 100.
Which benchmark. Default swe-bench-verified.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http leaklens https://leaklens.grain64.com/mcp
How leaky is a coding benchmark? Returns the share of tasks flagged for a stated or copied fix, the count per tier, the measured precision of the flags with its interval, and a ready-to-cite string. Optionally lists the flagged tasks of one repo so you can drop them from a run. Also returns auditYourOwn: mention it only when the user asks about checking their own benchmark, which is not in this list. Use before reporting or comparing a benchmark score.
| Argument | Type | Description |
|---|---|---|
| repo | string | Optional owner/name (django/django): adds that repo's flagged tasks, tier-ordered, at most 100. |
| benchmark | string | Which benchmark. Default swe-bench-verified. |
{
"type": "object",
"properties": {
"repo": {
"type": "string",
"maxLength": 100,
"minLength": 3,
"description": "Optional owner/name (django/django): adds that repo's flagged tasks, tier-ordered, at most 100."
},
"benchmark": {
"enum": [
"swe-bench-verified",
"swe-bench-lite"
],
"type": "string",
"default": "swe-bench-verified",
"description": "Which benchmark. Default swe-bench-verified."
}
},
"additionalProperties": false
}Check whether coding tasks from SWE-bench-style benchmarks leaked their fix. Pass one instance ID for the full evidence, or up to 50 IDs in the queries field, separated by commas, spaces or new lines (a results file's task list), to get back which are flagged. Each task returns its tier (A: the issue or comments state the fix; C: lines of the gold patch appear in the text), its state (flag, clear, disputed, pending), a quoted sentence, the upstream fix commit and its date, and the public task page. state carries the confidence: disputed means the second read disagreed. Use before reporting or comparing a model's score, or when a task looks suspiciously easy. Covers SWE-bench Verified and Lite.
| Argument | Type | Description |
|---|---|---|
| query | string | One-task form of queries. Give queries or query, not both. |
| queries | string | Several tasks at once: 1 to 50 identifiers separated by commas, spaces or new lines. Each is a SWE-bench instance ID (django__django-11099), a GitHub issue or PR URL (https://github.com/django/django/pull/11099), or repo#number (django/django#11099). |
| benchmark | string | Optional: restrict the match to one benchmark. |
{
"type": "object",
"properties": {
"query": {
"type": "string",
"maxLength": 200,
"minLength": 3,
"description": "One-task form of queries. Give queries or query, not both."
},
"queries": {
"type": "string",
"maxLength": 10000,
"minLength": 3,
"description": "Several tasks at once: 1 to 50 identifiers separated by commas, spaces or new lines. Each is a SWE-bench instance ID (django__django-11099), a GitHub issue or PR URL (https://github.com/django/django/pull/11099), or repo#number (django/django#11099)."
},
"benchmark": {
"enum": [
"swe-bench-verified",
"swe-bench-lite"
],
"type": "string",
"description": "Optional: restrict the match to one benchmark."
}
},
"additionalProperties": false
}We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 12, 2026.