Entroly
PYPI · ENTROLY · 2 COMPONENTS
Context Assurance for AI agents: context engineering, AI cost optimization, verification.
Available components
Last scanned (23 hours ago)
Scores higher than 89% of PyPI servers
Category AI and Machine Learning
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 Security100
- No malware found by supply-chain analysis.Pass
- No known CVEs affecting this package version or its production dependencies.Pass
- Runs hatchling.build at install time, a recognised build step with no custom scripting around it. View diagnostics → Pass
- 0 of 31 dependencies flagged as unhealthy. View diagnostics → Pass
Provenance & Transparency48
- Source repository is publicly reachable at the declared URL. View diagnostics → Pass
- Provenance check failed: no build-provenance attestation is published. See how to fix → View diagnostics → Fail
- Clear OSI-approved license (Apache-2.0).Pass
- Actively maintained (last published 3 days ago).Pass
- Publishes a security disclosure policy (SECURITY.md).Pass
Schema Quality & AI Usability83
- 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).Pass
- AI-judged instruction clarity (good).Pass
- Context-footprint check failed: tool/resource definitions use about 2721 tokens (~113/item across 24 items; 22 tools + 2 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 Management100
- No destabilizing schema changes in the last 30 days.Pass
Tool Coverage100
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 100% of tool parameters carry a description.Pass
- Structured output schemas are declared (100% of tools); any adoption earns full credit.Pass
Tool Safety100
- No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
- We read all 22 captured tool definition(s), and no name or description among them implies an irreversible operation.Pass
- An AI judge read all 24 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
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.
- 3 Nov 26 +15
- Malware scan: unverified → pass ▲ security
- 2 Nov 26 −15
- Malware scan: pass → unverified ▼ security
- Schema quality: 126 → 113 ▲ functional
- Schema quality: excellent → good functional
- Package version: 1.0.86 → 1.0.88 functional
- 1 Nov 26 +1
- Stability: 0.97 → pass security
- 31 Oct 26 −1
- Stability: pass → 0.97 functional
- 29 Oct 26 +15
- Malware scan: unverified → pass ▲ security
- 26 Oct 26 +13
- Stability: fail → pass ▲ security
- 25 Oct 26 −15
- Malware scan: pass → unverified ▼ security
- 23 Oct 26 +15
- Malware scan: unverified → 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 4 Nov 2026 · Analysed pypi/entroly@1.0.88
Provenance No attestation
The registry publishes no build provenance for this version, so there is nothing to verify.
| Result | No attestation |
|---|---|
| Ecosystem | pypi |
| Reason | No attestation published |
Background: How many MCP packages publish verified provenance →
Install scripts 1 script
| Hook | Tier | Command |
|---|---|---|
| build_backend | allowlisted | hatchling.build |
Background: Why install scripts are a supply-chain risk →
Dependencies 31 packages
| Packages resolved | 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 →
checkpoint_state ~133
Persist Entroly task state and explicit continuation metadata locally. Use before a planned handoff, restart, or context loss; use resume_state to restore it. This writes a local checkpoint, does not edit source files, and makes no network call.
| Name | Type | Req | Description |
|---|---|---|---|
| current_step | string | – | Last completed step or next action for continuation. |
| decisions | – | – | Important decisions that a resumed agent must preserve. |
| modified_files | – | – | Workspace paths already changed during the task. |
| project | string | – | Optional project path or identifier used to scope recovery. |
| task_description | string | – | Concise description used to identify the checkpoint later. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
create_context_receipt ~163
Create an auditable context-selection receipt from caller-supplied documents. Use this when you need selected and omitted chunks, fingerprints, warnings, and recovery metadata. It does not read workspace files or call a model. recoverable=true writes a local content-addressed recovery bundle.
| Name | Type | Req | Description |
|---|---|---|---|
| chunk_tokens | integer | – | Target size of each source chunk in estimated tokens. |
| documents_json | string | yes | JSON object, pairs, or objects that map source labels to document text. |
| overlap_tokens | integer | – | Estimated tokens repeated between adjacent chunks. |
| query | string | yes | Task used to select relevant document chunks. |
| recoverable | boolean | – | Persist omitted text locally for verified later recovery. |
| token_budget | integer | – | Maximum estimated tokens selected into the receipt. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
get_session_stats ~56
Read current Entroly session counters, selection statistics, checkpoint status, and locally estimated token or cost fields. Use for diagnostics after work has run. Estimates are not provider billing evidence. This is read-only and makes no network call.
Input schema present but exposes no named parameters.
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
locate_evidence ~232
Locate exact passages and sentence offsets in caller-supplied text. Use this when wording may differ from the query but the returned evidence must remain extractive and auditable. The default lexical ranker is deterministic. ENTROLY_SEMANTIC_MODEL_PATH may name an existing local sentence-transformer directory; remote downloads are refused. Scores rank supplied spans and do not prove relevance, completeness, factual correctness, or task success. This may write an exact local recovery entry and makes no network call.
| Name | Type | Req | Description |
|---|---|---|---|
| calibration_id | string | – | External calibration artifact identity for the score floor, if one exists. |
| max_matches | integer | – | Maximum number of exact source passages to return. |
| min_score | number | – | Optional score floor from 0 to 1; -1 disables it. |
| passage_mode | string | – | Source segmentation: auto, paragraph, or line. |
| query | string | yes | Natural-language description of the evidence to locate. |
| source_id | string | – | Provenance label copied into the receipt. |
| text | string | yes | Caller-supplied source text whose exact spans may be returned. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
optimize_context ~99
Select a token-budgeted context bundle from stored fragments for one task. Use this after fragments have been indexed. Unlike recall_relevant, it optimizes the whole bundle against token_budget; unlike read_source_file, it does not read a requested path. It updates local access statistics and makes no network call.
| Name | Type | Req | Description |
|---|---|---|---|
| query | string | – | Task description used to score fragment relevance. |
| token_budget | integer | – | Maximum estimated tokens in the selected context bundle. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
read_source_file ~254
Read one current workspace file at an exact or relevance-based resolution. Use this for a known path; use recall_relevant to search stored fragments and retrieve_context for a CCR handle. resolution=full or a line range returns exact source text. This is a read-only inspection tool: it cannot submit findings or a review verdict. It reads only within the configured project root and makes no network call.
| Name | Type | Req | Description |
|---|---|---|---|
| budget | integer | – | Target output size in estimated tokens for automatic resolution. |
| file_path | string | yes | Workspace-relative path constrained to the configured project root. |
| fresh | boolean | – | Return content again instead of a same-session repeat-delivery handle. |
| line_end | integer | – | Last line of an exact inclusive range; pair with line_start. |
| line_start | integer | – | First line of an exact inclusive range; pair with line_end. |
| previous_source | string | – | Required baseline text when resolution is diff. |
| query | string | – | Optional task terms used for automatic relevance-based resolution. |
| read_scope | string | – | Optional cache-isolation key for agents sharing one MCP connection. |
| resolution | string | – | One of full, medium, diff, structure, low, or empty for automatic selection. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
recall_relevant ~115
Search Entroly's stored fragment index for the best task matches. Use this for a ranked lookup when no token-bounded bundle is needed; use optimize_context to assemble a complete budgeted context. The default returns compact pointers, while full returns stored bodies. This reads local state and makes no network call.
| Name | Type | Req | Description |
|---|---|---|---|
| full | boolean | – | Return complete stored bodies instead of compact pointers. |
| query | string | yes | Search terms derived from the current task. |
| top_k | integer | – | Maximum number of ranked fragments to return. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
record_test_result ~128
Attach an already-observed test pass or failure to an optimization trace. Use only after a real test command ran; this tool does not run tests, inspect CI, or independently validate the claim. It appends a local evidence event and makes no network call.
| Name | Type | Req | Description |
|---|---|---|---|
| details | string | – | Optional bounded summary of the observed test execution. |
| passed | boolean | yes | True only when every test required by the named suite passed. |
| request_id | string | yes | Trace identifier returned by the relevant optimize_context call. |
| suite | string | – | Optional test-suite name such as pytest or cargo test. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
recover_receipt_omission ~102
Recover omitted text from a recoverable Context Receipt and verify its stored hashes. Use this only with receipt JSON created using recoverable=true. Use retrieve_context for a CCR handle. This reads the local recovery store, does not modify source files, and makes no network call.
| Name | Type | Req | Description |
|---|---|---|---|
| chunk_id | string | – | Specific omitted chunk to recover; empty recovers every omitted chunk. |
| receipt_json | string | yes | Complete JSON returned by create_context_receipt. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
remember_fragment ~153
Store caller-supplied text in Entroly's local fragment index. Use this for transient text or tool output that is not already a workspace file; use read_source_file for a file and recall_relevant to search stored fragments. The context firewall may reject unsafe content. This writes local Entroly state and makes no network call.
| Name | Type | Req | Description |
|---|---|---|---|
| content | string | yes | Text to scan and store in the local fragment index. |
| is_pinned | boolean | – | Prioritize the fragment within the bounded pinned reserve. |
| source | string | – | Optional provenance label such as file:utils.py or tool:pytest. |
| token_count | integer | – | Known token count; use 0 to let Entroly estimate it. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
resume_state ~97
Restore a local Entroly checkpoint into the current engine. Use after a restart or context loss; use checkpoint_state to create checkpoints and recall_relevant for read-only fragment search. This changes in-memory Entroly state, does not edit source files, and makes no network call.
| Name | Type | Req | Description |
|---|---|---|---|
| project | string | – | Optional project path or identifier used to narrow matching. |
| query | string | – | Optional task terms; empty selects the latest checkpoint. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
retrieve_context ~91
Recover exact content from a CCR retrieval handle or materialized source reference. Use this when compressed context includes a ccr: handle. Use read_source_file for a current workspace file and recover_receipt_omission for an omitted receipt chunk. This is a read-only local lookup and makes no network call.
| Name | Type | Req | Description |
|---|---|---|---|
| source_or_handle | string | – | CCR handle or previously materialized source path; empty lists available entries. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
verify_response ~117
Estimate whether an AI response is grounded in caller-supplied context using Entroly's local verification signals. Use after generation to identify claims requiring review. A pass is a risk estimate, not formal proof or external fact-checking. The tool is read-only and makes no model or network call.
| Name | Type | Req | Description |
|---|---|---|---|
| context | string | – | Source evidence that was actually available to the AI. |
| prompt | string | – | Original user request used only to calibrate the local analysis. |
| response | string | yes | AI-generated text whose claims should be checked. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
work_acknowledge_recovery ~51
Acknowledge the exact recovered state before claiming follow-up work.
| Name | Type | Req | Description |
|---|---|---|---|
| project | string | – | Optional path within the configured repository root. |
| token | string | yes | Token returned by work_resume. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
work_capabilities ~41
Check the active repository identity and local Work Graph review support. No model or network call.
| Name | Type | Req | Description |
|---|---|---|---|
| project | string | – | Optional path within the configured repository root. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
work_claim ~134
Record an explicit agent task and advisory scope lease in the local Work Graph.
| Name | Type | Req | Description |
|---|---|---|---|
| agent_id | string | yes | Caller-supplied agent identifier; not an identity attestation. |
| project | string | – | Optional path within the configured repository root. |
| scope_paths | – | – | Advisory repository-relative paths in scope. |
| scope_symbols | – | – | Advisory code symbols in scope. |
| session_id | string | – | Optional caller session identifier. |
| task_id | string | – | Stable task identifier supplied by the caller. |
| task_title | string | yes | Short title of the task being claimed. |
| ttl_seconds | number | – | Lease duration in seconds. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
work_handoff ~84
Seal a graph-bound handoff and save a compact local inbox receipt for another agent.
| Name | Type | Req | Description |
|---|---|---|---|
| from_agent | string | yes | Caller-supplied writer identifier. |
| project | string | – | Optional path within the configured repository root. |
| to_agent | string | yes | Caller-supplied reviewer identifier. |
| workstream_id | string | yes | Existing workstream identifier from work_state. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
work_pending_handoffs ~54
List local handoffs for an agent and verify each against current repository state.
| Name | Type | Req | Description |
|---|---|---|---|
| project | string | – | Optional path within the configured repository root. |
| to_agent | string | yes | Reviewer identifier used by the writer. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
work_record_review ~184
Submit a caller-authored review verdict, findings, and observed checks against a verified handoff receipt. This writes a local review record after reading bounded files to bind their digests; it does not return source text or perform the review for the caller. The verdict is an agent statement, not a correctness proof.
| Name | Type | Req | Description |
|---|---|---|---|
| checks | – | – | Optional observed check names and exit codes; caller statements only. |
| findings | array | yes | Caller-supplied path, line, severity, and message records. |
| handoff | object | yes | Exact receipt returned by work_handoff or work_pending_handoffs. |
| project | string | – | Optional path within the configured repository root. |
| reviewer_id | string | yes | Reviewer identifier matching the handoff target; self asserted. |
| summary | string | yes | Short reviewer conclusion. |
| verdict | string | yes | approve, changes_requested, or comment. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
work_resume ~87
Recover durable work evidence, refresh the local graph, and arm an acknowledgement gate.
| Name | Type | Req | Description |
|---|---|---|---|
| max_evidence | integer | – | Maximum evidence references returned. |
| project | string | – | Optional path within the configured repository root. |
| to_agent | string | – | Optional receiving agent for a no-handoff proof. |
| workstream_id | string | – | Workstream to recover; empty chooses resumable work. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
work_reviews ~47
Read recorded review statements and whether their source state is current.
| Name | Type | Req | Description |
|---|---|---|---|
| project | string | – | Optional path within the configured repository root. |
| workstream_id | string | – | Optional workstream filter. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
work_state ~58
Inspect shared Work Graph state. First use may start local takeover and append repository evidence.
| Name | Type | Req | Description |
|---|---|---|---|
| now_ms | integer | – | Optional time for lease status; zero uses the current time. |
| project | string | – | Optional path within the configured repository root. |
| Name | Type | Req | Description |
|---|---|---|---|
| result | string | yes | – |
No examples provided.
How do I install the Entroly MCP server?
Entroly runs locally as a PyPI package, launched with uvx entroly. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
pypi · entroly
claude mcp add juyterman1000-entroly -- uvx entroly
What is the Entroly MCP server?
Entroly is an MCP server listed in the public MCP registry as io.github.juyterman1000/entroly. Context Assurance for AI agents: context engineering, AI cost optimization, verification. This page covers its PyPI package (entroly).
Is the Entroly MCP server safe to use?
Entroly scores 87 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 4 November 2026. 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 Entroly MCP server expose?
Entroly exposes 22 tools: remember_fragment, optimize_context, recall_relevant, record_test_result, create_context_receipt, and 17 more. Their descriptions and schemas cost roughly 2,480 tokens of context every time the server is loaded.
Is the Entroly MCP server still maintained?
Entroly is still listed as active in the MCP registry. We last reached this channel on 4 November 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 Entroly MCP server under?
Entroly declares the Apache-2.0 licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.