DeepThonk
NPM · DEEPTHONK · SCANNED SEP 20
Provider-neutral OpenDeepThink reasoning server: generate, judge, rank, mutate.
Available components
How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →
Supply Chain Security98
- No malware found by supply-chain analysis.Pass
- No known CVEs affecting this package version or its production dependencies.Pass
- No install/post-install scripts declared.Pass
- 31 of 100 dependencies flagged as unhealthy. View diagnostics → Partial
Provenance & Transparency100
- Source repository is publicly reachable at the declared URL. View diagnostics → Pass
- Cryptographically verified build provenance (signed, bound to linxule/deepthonk). View diagnostics → Pass
- Clear OSI-approved license (MIT).Pass
- Actively maintained (last published 5 days ago).Pass
- Publishes a security disclosure policy (SECURITY.md).Pass
Schema Quality & AI Usability65
- 80% of prompts and resources have a non-trivial description (not blank, and not just the item's name).Partial
- AI-judged instruction clarity (good).Pass
- Context-footprint check failed: tool/resource definitions use about 3360 tokens (~186/item across 18 items; 17 tools + 1 resources), over budget; trim descriptions and params. See how to fix → Fail
- Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management83
- Stability observed for 25 of 30 days with no destabilising changes; credit accrues until the full window elapses.Partial
Tool Coverage95
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 82% of tool parameters carry a description.Partial
- Structured output schemas are declared (100% of tools); any adoption earns full credit.Pass
Tool Safety75
- No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
- 0 of 1 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation; "deepthonk.profile_delete" implies "delete" and declares no destructiveHint at all, which the MCP spec reads as destructive by default. See how to fix → Fail
- An AI judge read all 18 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
- Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.Pass
How do I install the DeepThonk MCP server?
DeepThonk runs locally as an npm package, launched with npx -y deepthonk. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
npm · deepthonk
claude mcp add linxule-deepthonk -- npx -y deepthonk
{
"mcpServers": {
"linxule-deepthonk": {
"command": "npx",
"args": [
"-y",
"deepthonk"
]
}
}
} {
"servers": {
"linxule-deepthonk": {
"command": "npx",
"args": [
"-y",
"deepthonk"
]
}
}
} codex mcp add linxule-deepthonk -- npx -y deepthonk
{
"$schema": "https://opencode.ai/config.json",
"mcp": {
"linxule-deepthonk": {
"type": "local",
"command": [
"npx",
"-y",
"deepthonk"
],
"enabled": true
}
}
} openclaw mcp add linxule-deepthonk --command npx --arg -y --arg deepthonk
mcp_servers:
linxule-deepthonk:
command: "npx"
args: ["-y", "deepthonk"] {
"McpServers": {
"linxule-deepthonk": {
"Transport": "stdio",
"Command": "npx",
"Arguments": [
"-y",
"deepthonk"
]
}
}
} assistant mcp add linxule-deepthonk -t stdio -c npx -a -y deepthonk
{
"mcpServers": {
"linxule-deepthonk": {
"command": "npx",
"args": [
"-y",
"deepthonk"
]
}
}
} Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.
- 20 Sept 26 −2
- Stability: pass → 0.83 functional
- 19 Sept 26 0
- Stability: 0.97 → pass security
- 18 Sept 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 93 to 97. That category is still filling its 30-day observation window: 28 days of observed history at the previous scan, 29 at this one. The score rises as the window fills, whether or not the server changes.
- 16 Sept 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 87 to 90. That category is still filling its 30-day observation window: 26 days of observed history at the previous scan, 27 at this one. The score rises as the window fills, whether or not the server changes.
- 15 Sept 26 +26
- Malware scan: unverified → pass ▲ security
- Known CVEs: unverified → pass ▲ security
- Dependency health: unverified → 0.87 ▲ functional
- 14 Sept 26 −25
- Stability: 0.80 → unverified ▼ security
- Malware scan: pass → unverified ▼ security
- Known CVEs: pass → unverified ▼ security
- Tool safety: pass → unverified ▼ security
- Capabilities: pass → unverified ▼ functional
- Dependency health: 0.87 → unverified ▼ functional
- Tool coverage: 100 → unverified ▼ functional
- Schema quality: 80 → unverified ▼ functional
- Package version: 0.3.0 → 0.3.1 functional
- 13 Sept 26 −3
- Stability: pass → 0.80 functional
- 12 Sept 26 0
- Stability: 0.97 → pass security
Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.
Captured 20 Sept 2026 · Analysed npm/deepthonk@0.3.1
Provenance Verified
A signed build attestation was found and verified, binding this exact artifact to the source repository it claims to come from.
| Result | Verified |
|---|---|
| Ecosystem | npm |
| Reason | Verified |
| Discovered via | Registry attestation endpoint |
| Source repo | linxule/deepthonk |
| Certificate issuer | https://token.actions.githubusercontent.com |
| Certificate SAN | https://github.com/linxule/deepthonk/.github/workflows/publish.yml@refs/tags/v0.3.1 |
| Rekor log index | 2830090179 |
| Predicate type | https://slsa.dev/provenance/v1 |
| Subject digest | sha512:75ef0e63219f429e98f229d6349a6e0fdfa41da23d20db05fd86b9550d1231a803a9e47aad5ecb828ec1aa4950de15ecba27266cb2ebc5ddd333bb657 |
Background: How many MCP packages publish verified provenance →
Dependencies 100 packages
| Packages resolved | 100 |
|---|---|
| Stale | 31 |
| Tree resolution | Complete |
Background: SBOMs and build attestations, explained →
The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →
deepthonk.cancel Cancel DeepThonk Job ~62
Request cancellation for a running job by writing cancel.json in the run directory.
| Name | Type | Req | Description |
|---|---|---|---|
| job_id | string | – | Job ID returned by deepthonk.start. |
| run_dir | string | – | Run directory returned by deepthonk.start or deepthonk.run. |
| Name | Type | Req | Description |
|---|---|---|---|
| cancel_requested | boolean | yes | – |
| run_dir | string | – | – |
No examples provided.
deepthonk.export Export Run ~32
Export a run summary or trace.
| Name | Type | Req | Description |
|---|---|---|---|
| format | string | – | – |
| run_dir | string | yes | – |
Structured output declared, but exposes no named fields.
No examples provided.
deepthonk.lock_inspect Inspect DeepThonk Run Lock ~40
Inspect run.lock and return its exact fingerprint without mutating it.
| Name | Type | Req | Description |
|---|---|---|---|
| run_dir | string | yes | Run directory containing the lock to inspect. |
| Name | Type | Req | Description |
|---|---|---|---|
| fingerprint | string | – | – |
| lock | object | – | – |
| run_dir | string | yes | – |
| sameHost | boolean | – | – |
| state | string | yes | – |
| workerAlive | boolean | – | – |
No examples provided.
deepthonk.lock_reclaim Reclaim DeepThonk Run Lock ~59
Reclaim run.lock only when its current bytes match the inspected fingerprint exactly.
| Name | Type | Req | Description |
|---|---|---|---|
| fingerprint | string | yes | Exact fingerprint returned by deepthonk.lock_inspect. |
| run_dir | string | yes | Run directory containing the lock to inspect. |
| Name | Type | Req | Description |
|---|---|---|---|
| fingerprint | string | yes | – |
| reclaimed | boolean | yes | – |
| run_dir | string | yes | – |
No examples provided.
deepthonk.mutate Mutate Candidate ~437
Mutate one supplied candidate with critique.
| Name | Type | Req | Description |
|---|---|---|---|
| api_key_env | string | – | Environment variable that contains the API key. |
| base_url | string | – | OpenAI-compatible base URL, ending at /v1. |
| candidate | string | yes | Candidate text to mutate. |
| config_path | string | – | Optional DeepThonk YAML config path. |
| critique | string | – | Critique or guidance for mutation. |
| finalizer_model | string | – | Optional model used to polish the winning candidate. |
| generator_model | string | – | Model used for initial candidate generation. |
| judge_model | string | – | Model used for pairwise judging. |
| model_output_tokens | object | – | Per-role model output token caps for one-shot operations. |
| mutate_temperature | number | – | Temperature for mutation. |
| mutator_model | string | – | Model used for critique-guided mutation. |
| profile | string | – | Profile used for one-shot defaults. |
| profile_name | string | – | Saved profile bundle name; loads ~/.config/deepthonk/profiles/<name>.yaml. Mutually exclusive with config_path. |
| prompt_style | string | – | – |
| prompts | object | – | – |
| provider | string | – | Provider label. Built-ins include fake, deepseek, openrouter, and openai-compatible; custom labels require base_url. |
| provider_max_concurrency | integer | – | Process-shared maximum concurrent calls to this provider route. |
| request_timeout_ms | integer | – | Logical provider-call deadline in milliseconds. |
| rubric | string | – | Optional rubric text. |
| sampling_cost_priority | number | – | MCP Sampling model preference for lower cost, from 0 to 1. |
| sampling_intelligence_priority | number | – | MCP Sampling model preference for intelligence, from 0 to 1. |
| sampling_model_hints | array | – | Optional MCP Sampling model hints. |
| sampling_speed_priority | number | – | MCP Sampling model preference for speed, from 0 to 1. |
| supports_json_mode | boolean | – | Whether the base OpenAI-compatible provider supports response_format JSON mode. |
| task | string | yes | Task text the candidate answers. |
| Name | Type | Req | Description |
|---|---|---|---|
| model | string | – | – |
| mutated | string | yes | – |
| provider | string | – | – |
| usage | object | – | – |
No examples provided.
deepthonk.plan Plan DeepThonk Budget ~103
Estimate model calls for a DeepThonk profile.
| Name | Type | Req | Description |
|---|---|---|---|
| config_path | string | – | Optional DeepThonk YAML config path. |
| k | integer | – | – |
| m | integer | – | – |
| n | integer | – | – |
| profile | string | – | – |
| profile_name | string | – | Saved profile bundle name; loads ~/.config/deepthonk/profiles/<name>.yaml. Mutually exclusive with config_path. |
| t | integer | – | – |
Structured output declared, but exposes no named fields.
No examples provided.
deepthonk.profile_delete Delete DeepThonk Profile ~24
Delete a saved named profile.
| Name | Type | Req | Description |
|---|---|---|---|
| name | string | yes | – |
| Name | Type | Req | Description |
|---|---|---|---|
| deleted | string | yes | – |
No examples provided.
deepthonk.profile_list List DeepThonk Profiles ~16
List saved named profiles.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| profiles | array | yes | – |
No examples provided.
deepthonk.profile_save Save DeepThonk Profile ~119
Save a reusable named profile bundle.
| Name | Type | Req | Description |
|---|---|---|---|
| algorithm | object | – | – |
| api_key_env | string | – | – |
| base_url | string | – | – |
| budget | – | – | – |
| concurrency | – | – | – |
| force | boolean | – | – |
| models | object | – | – |
| name | string | yes | – |
| output | – | – | – |
| profile | string | – | – |
| prompt_style | string | – | – |
| prompts | object | – | – |
| provider | string | – | – |
| providers | object | – | – |
| retry | – | – | – |
| Name | Type | Req | Description |
|---|---|---|---|
| path | string | yes | – |
No examples provided.
deepthonk.profile_show Show DeepThonk Profile ~34
Show a saved named profile; manually edited secret-shaped values are rejected on load.
| Name | Type | Req | Description |
|---|---|---|---|
| name | string | yes | – |
| Name | Type | Req | Description |
|---|---|---|---|
| profile | object | yes | – |
No examples provided.
deepthonk.rank Rank Candidates ~481
Rank supplied candidates with pairwise judging and Bradley-Terry aggregation.
| Name | Type | Req | Description |
|---|---|---|---|
| api_key_env | string | – | Environment variable that contains the API key. |
| base_url | string | – | OpenAI-compatible base URL, ending at /v1. |
| candidates | array | yes | Candidate texts or {id, content} objects. |
| concurrency | integer | – | Maximum concurrent pairwise comparisons. |
| config_path | string | – | Optional DeepThonk YAML config path. |
| finalizer_model | string | – | Optional model used to polish the winning candidate. |
| generator_model | string | – | Model used for initial candidate generation. |
| judge_model | string | – | Model used for pairwise judging. |
| judge_temperature | number | – | Temperature for pairwise judging. |
| lambda | number | – | Bradley-Terry L2 regularization. |
| model_output_tokens | object | – | Per-role model output token caps for one-shot operations. |
| mutator_model | string | – | Model used for critique-guided mutation. |
| profile | string | – | Profile used for one-shot defaults. |
| profile_name | string | – | Saved profile bundle name; loads ~/.config/deepthonk/profiles/<name>.yaml. Mutually exclusive with config_path. |
| prompt_style | string | – | – |
| prompts | object | – | – |
| provider | string | – | Provider label. Built-ins include fake, deepseek, openrouter, and openai-compatible; custom labels require base_url. |
| provider_max_concurrency | integer | – | Process-shared maximum concurrent calls to this provider route. |
| rank | object | – | Pair scheduling mode, degree, seed, and logical-call limit. |
| request_timeout_ms | integer | – | Logical provider-call deadline in milliseconds. |
| rubric | string | – | Optional judging rubric text. |
| sampling_cost_priority | number | – | MCP Sampling model preference for lower cost, from 0 to 1. |
| sampling_intelligence_priority | number | – | MCP Sampling model preference for intelligence, from 0 to 1. |
| sampling_model_hints | array | – | Optional MCP Sampling model hints. |
| sampling_speed_priority | number | – | MCP Sampling model preference for speed, from 0 to 1. |
| supports_json_mode | boolean | – | Whether the base OpenAI-compatible provider supports response_format JSON mode. |
| task | string | yes | Task text the candidates answer. |
| Name | Type | Req | Description |
|---|---|---|---|
| comparisons | array | yes | – |
| scores | array | yes | – |
No examples provided.
deepthonk.repair_budget Repair Legacy Budget Config ~42
Replace only legacy [redacted] numeric budget fields with explicit original values.
| Name | Type | Req | Description |
|---|---|---|---|
| replacements | object | yes | – |
| run_dir | string | yes | – |
| Name | Type | Req | Description |
|---|---|---|---|
| repaired | array | yes | – |
| run_dir | string | yes | – |
No examples provided.
deepthonk.result DeepThonk Job Result ~60
Return final summary/resource URIs when a background job is complete.
| Name | Type | Req | Description |
|---|---|---|---|
| job_id | string | – | Job ID returned by deepthonk.start. |
| run_dir | string | – | Run directory returned by deepthonk.start or deepthonk.run. |
| Name | Type | Req | Description |
|---|---|---|---|
| complete | boolean | – | – |
| run_dir | string | – | – |
| run_id | string | – | – |
No examples provided.
deepthonk.resume Resume Run ~41
Detect resume state (default) or replay an interrupted run with continue: true.
| Name | Type | Req | Description |
|---|---|---|---|
| continue | boolean | – | – |
| run_dir | string | yes | – |
| Name | Type | Req | Description |
|---|---|---|---|
| generation | – | – | – |
| message | string | yes | – |
| phase | string | – | – |
| run_id | string | – | – |
| safe_to_continue | boolean | – | – |
| status | string | yes | – |
No examples provided.
deepthonk.run Run DeepThonk ~870
Run DeepThonk through the shared core engine.
| Name | Type | Req | Description |
|---|---|---|---|
| api_key_env | string | – | Environment variable that contains the API key. |
| base_url | string | – | OpenAI-compatible base URL, ending at /v1. |
| concurrency | object | – | Per-phase concurrency overrides. |
| config_path | string | – | Optional DeepThonk YAML config path, such as ~/.config/deepthonk/config.yaml. |
| critique_limits | object | – | Critique aggregation bounds. |
| finalizer_model | string | – | Optional model used to polish the winning candidate. |
| generator_model | string | – | Model used for initial candidate generation. |
| include_prompts | boolean | – | Store rendered prompts in candidate/comparison metadata. |
| include_raw_model_outputs | boolean | – | Store raw provider responses in trace metadata. |
| judge_model | string | – | Model used for pairwise judging. |
| judge_temperature | number | – | Temperature for pairwise judging. |
| k | integer | – | Comparisons per candidate per mutation-generation round. |
| lambda | number | – | Bradley-Terry L2 regularization. Defaults to 0.01. |
| m | integer | – | Comparisons per candidate in the final dense ranking round. Mutation count per generation is n - ceil(n/4), not m. |
| max_calls | integer | – | Maximum logical model invocations, including failed calls and invalid-JSON retries; internal HTTP retries are reported separately. |
| max_input_tokens | integer | – | Maximum recorded input tokens before the run stops at a phase boundary. |
| max_output_tokens | integer | – | Maximum recorded output tokens before the run stops at a phase boundary. |
| max_usd | number | – | Maximum estimated USD spend; requires matching budget.prices in config. |
| model_output_tokens | object | – | Per-role model output token caps. |
| mutate_temperature | number | – | Temperature for critique-guided mutation. |
| mutator_model | string | – | Model used for critique-guided mutation. |
| n | integer | – | Population size override. Defaults to the profile's n. |
| profile | string | – | Run profile. quick is safest for smoke tests; paper plans 285 calls. |
| profile_name | string | – | Saved profile bundle name; loads ~/.config/deepthonk/profiles/<name>.yaml. Mutually exclusive with config_path. |
| prompt_style | string | – | Prompt template style. Defaults to paper-programming for the paper profile, general otherwise. |
| prompts | object | – | Optional per-phase prompt template overrides. Templates use {task}, {rubric}, {candidate}, {candidateA}, {candidateB}, {critique} placeholders. Use {{ and }} to escape literal braces. See docs/custom… |
| provider | string | – | Provider label. Built-ins include fake, deepseek, openrouter, and openai-compatible; custom labels require base_url. |
| provider_max_concurrency | integer | – | Process-shared maximum concurrent calls to this provider route. |
| rank | object | – | Final ranking mode, degree, seed, and logical-call limit. |
| request_timeout_ms | integer | – | Logical provider-call deadline in milliseconds, including body reads and retry waits. |
| rubric | string | – | Optional judging rubric text. |
| run_dir | string | – | Directory for trace files. |
| run_id | string | – | Optional stable caller-supplied run ID. |
| sample_temperature | number | – | Temperature for initial candidate generation. |
| sampling_cost_priority | number | – | MCP Sampling model preference for lower cost, from 0 to 1. |
| sampling_intelligence_priority | number | – | MCP Sampling model preference for intelligence, from 0 to 1. |
| sampling_model_hints | array | – | Optional MCP Sampling model hints. Hints guide host model choice but do not enforce it. |
| sampling_speed_priority | number | – | MCP Sampling model preference for speed, from 0 to 1. |
| seed | integer | – | Deterministic seed for pair ordering and IDs. |
| supports_json_mode | boolean | – | Whether the base OpenAI-compatible provider supports response_format JSON mode. |
| t | integer | – | Number of mutation generations (t=0 disables mutation rounds). |
| task | string | yes | Task text to solve. MCP tools do not read task files. |
| Name | Type | Req | Description |
|---|---|---|---|
| final_answer | string | yes | – |
| run_dir | string | yes | – |
| run_id | string | yes | – |
| summary_resource | string | yes | – |
| trace_resource | string | yes | – |
| winner_id | string | yes | – |
No examples provided.
deepthonk.start Start DeepThonk Job ~875
Start DeepThonk in the background and persist status in the run directory.
| Name | Type | Req | Description |
|---|---|---|---|
| api_key_env | string | – | Environment variable that contains the API key. |
| base_url | string | – | OpenAI-compatible base URL, ending at /v1. |
| concurrency | object | – | Per-phase concurrency overrides. |
| config_path | string | – | Optional DeepThonk YAML config path, such as ~/.config/deepthonk/config.yaml. |
| critique_limits | object | – | Critique aggregation bounds. |
| finalizer_model | string | – | Optional model used to polish the winning candidate. |
| generator_model | string | – | Model used for initial candidate generation. |
| include_prompts | boolean | – | Store rendered prompts in candidate/comparison metadata. |
| include_raw_model_outputs | boolean | – | Store raw provider responses in trace metadata. |
| judge_model | string | – | Model used for pairwise judging. |
| judge_temperature | number | – | Temperature for pairwise judging. |
| k | integer | – | Comparisons per candidate per mutation-generation round. |
| lambda | number | – | Bradley-Terry L2 regularization. Defaults to 0.01. |
| m | integer | – | Comparisons per candidate in the final dense ranking round. Mutation count per generation is n - ceil(n/4), not m. |
| max_calls | integer | – | Maximum logical model invocations, including failed calls and invalid-JSON retries; internal HTTP retries are reported separately. |
| max_input_tokens | integer | – | Maximum recorded input tokens before the run stops at a phase boundary. |
| max_output_tokens | integer | – | Maximum recorded output tokens before the run stops at a phase boundary. |
| max_usd | number | – | Maximum estimated USD spend; requires matching budget.prices in config. |
| model_output_tokens | object | – | Per-role model output token caps. |
| mutate_temperature | number | – | Temperature for critique-guided mutation. |
| mutator_model | string | – | Model used for critique-guided mutation. |
| n | integer | – | Population size override. Defaults to the profile's n. |
| profile | string | – | Run profile. quick is safest for smoke tests; paper plans 285 calls. |
| profile_name | string | – | Saved profile bundle name; loads ~/.config/deepthonk/profiles/<name>.yaml. Mutually exclusive with config_path. |
| prompt_style | string | – | Prompt template style. Defaults to paper-programming for the paper profile, general otherwise. |
| prompts | object | – | Optional per-phase prompt template overrides. Templates use {task}, {rubric}, {candidate}, {candidateA}, {candidateB}, {critique} placeholders. Use {{ and }} to escape literal braces. See docs/custom… |
| provider | string | – | Provider label. Built-ins include fake, deepseek, openrouter, and openai-compatible; custom labels require base_url. |
| provider_max_concurrency | integer | – | Process-shared maximum concurrent calls to this provider route. |
| rank | object | – | Final ranking mode, degree, seed, and logical-call limit. |
| request_timeout_ms | integer | – | Logical provider-call deadline in milliseconds, including body reads and retry waits. |
| rubric | string | – | Optional judging rubric text. |
| run_dir | string | – | Directory for trace files. |
| run_id | string | – | Optional stable caller-supplied run ID. |
| sample_temperature | number | – | Temperature for initial candidate generation. |
| sampling_cost_priority | number | – | MCP Sampling model preference for lower cost, from 0 to 1. |
| sampling_intelligence_priority | number | – | MCP Sampling model preference for intelligence, from 0 to 1. |
| sampling_model_hints | array | – | Optional MCP Sampling model hints. Hints guide host model choice but do not enforce it. |
| sampling_speed_priority | number | – | MCP Sampling model preference for speed, from 0 to 1. |
| seed | integer | – | Deterministic seed for pair ordering and IDs. |
| supports_json_mode | boolean | – | Whether the base OpenAI-compatible provider supports response_format JSON mode. |
| t | integer | – | Number of mutation generations (t=0 disables mutation rounds). |
| task | string | yes | Task text to solve. MCP tools do not read task files. |
| Name | Type | Req | Description |
|---|---|---|---|
| artifact_resources | object | yes | – |
| job_id | string | yes | – |
| result_resource | string | yes | – |
| run_dir | string | yes | – |
| state | string | yes | – |
| status_resource | string | yes | – |
No examples provided.
deepthonk.status DeepThonk Job Status ~57
Read persisted job/run status from a run directory.
| Name | Type | Req | Description |
|---|---|---|---|
| job_id | string | – | Job ID returned by deepthonk.start. |
| run_dir | string | – | Run directory returned by deepthonk.start or deepthonk.run. |
| Name | Type | Req | Description |
|---|---|---|---|
| error | object | – | – |
| generation | – | – | – |
| job_id | string | – | – |
| message | string | – | – |
| phase | string | – | – |
| run_dir | string | – | – |
| run_id | string | – | – |
| state | string | – | – |
| status | string | – | – |
| usage | object | – | – |
No examples provided.
What is the DeepThonk MCP server?
DeepThonk is an MCP server listed in the public MCP registry as io.github.linxule/deepthonk. Provider-neutral OpenDeepThink reasoning server: generate, judge, rank, mutate. This page covers its npm package (deepthonk).
Is the DeepThonk MCP server safe to use?
DeepThonk scores 89 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 20 September 2026. It declares no install or post-install scripts. Its build provenance is signed and verified. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.
What tools does the DeepThonk MCP server expose?
DeepThonk exposes 17 tools: deepthonk.plan, deepthonk.start, deepthonk.status, deepthonk.result, deepthonk.cancel, and 12 more. Their descriptions and schemas cost roughly 3,352 tokens of context every time the server is loaded.
Is the DeepThonk MCP server still maintained?
DeepThonk is still listed as active in the MCP registry. We last reached this channel on 20 September 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.
What licence is the DeepThonk MCP server under?
DeepThonk declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.