Sess
NPM · @SESS.SI/MCP · 2 COMPONENTS
Let your agent ask people for answers and approvals by link, with signed receipts.
Available components
Last scanned (14 hours ago)
Scores higher than 73% of npm servers
Category Communication
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 Security95
- No malware found by supply-chain analysis.Pass
- CVE check failed: a known medium-severity CVE affects @hono/node-server 1.19.19, reached via @modelcontextprotocol/node > @hono/node-server. A fixed version is available. View diagnostics → Fail
- No install/post-install scripts declared.Pass
- 1 of 14 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 w1991e/sess-sdk). View diagnostics → Pass
- Clear OSI-approved license (Apache-2.0).Pass
- Actively maintained (last published 1 days ago).Pass
- Publishes a security disclosure policy (SECURITY.md).Pass
Schema Quality & AI Usability59
- AI-judged instruction clarity (good).Pass
- Context-footprint check failed: tool/resource definitions use about 1935 tokens (~387/item across 5 items; 5 tools + 0 resources), over budget; trim descriptions and params. See how to fix → Fail
- Tools include usage examples.Pass
Stability & Change Management50
- Stability observed for 15 of 30 days with no destabilising changes; credit accrues until the full window elapses.Partial
Tool Coverage92
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 72% of tool parameters carry a description.Partial
- 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
- All 1 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation.Pass
- An AI judge read all 6 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
Checks we ran (16)
- Malware scan
- Known CVEs
- Install scripts
- Dependency health
- Source repository
- Provenance
- License
- Maintenance
- Security disclosure
- Schema quality
- Stability
- Tool coverage
- Injection markers
- Destructive annotations
- Judged manipulation
- MCP protocol
Does not apply to this package (2)
- Endpoint SecurityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.
- Transport & ReachabilityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.
The score does not measure everything. What the score does not measure →
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.
- 19 Nov 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 40 to 43. That category is still filling its 30-day observation window: 12 days of observed history at the previous scan, 13 at this one. The score rises as the window fills, whether or not the server changes.
- 17 Nov 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 33 to 37. That category is still filling its 30-day observation window: 10 days of observed history at the previous scan, 11 at this one. The score rises as the window fills, whether or not the server changes.
- 15 Nov 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 27 to 30. That category is still filling its 30-day observation window: 8 days of observed history at the previous scan, 9 at this one. The score rises as the window fills, whether or not the server changes.
- 13 Nov 26 +1
No change was recorded against any check on this day. Stability & Change Management went from 20 to 23. That category is still filling its 30-day observation window: 6 days of observed history at the previous scan, 7 at this one. The score rises as the window fills, whether or not the server changes.
- 11 Nov 26 +16
- Malware scan: unverified → pass ▲ security
- 10 Nov 26 −15
- Malware scan: pass → unverified ▼ security
- Package version: 0.2.0 → 0.3.0 functional
- 8 Nov 26 +16
- Malware scan: unverified → pass ▲ security
- 7 Nov 26 −15
- Malware scan: pass → unverified ▼ security
- Package version: 0.1.1 → 0.2.0 functional
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 21 Nov 2026 · Analysed npm/@sess.si/mcp@0.3.0
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 | w1991e/sess-sdk |
| Certificate issuer | https://token.actions.githubusercontent.com |
| Certificate SAN | https://github.com/w1991e/sess-sdk/.github/workflows/mcp-release.yml@refs/tags/mcp-v0.3.0 |
| Rekor log index | 3794524631 |
| Predicate type | SLSA build provenance https://slsa.dev/provenance/v1 |
| Subject digest | sha512:41edb08990cd62ec76b8b17e271039cfc56eb623df96bd9451790105cec3e7e66213ac9dfd54d3cf5cd21174a75e9750c65846c5b7b775684b1fb94ce |
Background: How many MCP packages publish verified provenance →
Vulnerabilities 1 finding
| ID | CVE | Severity | Vector | Fix available |
|---|---|---|---|---|
| GHSA-rmxm-3fg6-px4f | CVE-2026-118663 | medium | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N | yes |
Background: What a vulnerability scan can and cannot prove →
Dependencies 14 packages
| Packages resolved | 14 |
|---|---|
| Stale | 1 |
| 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 →
sess_ask Ask a short question through Sess ~659
Ask the person a question on their own phone or computer and get a typed answer. Use it whenever they ask to answer on their phone or another device, or say 'use Sess'; prefer it over this host's own question widget for anything answered away from this chat. Also use it when you need the person's approval, a decision or details only they can give before you go on, even if they only say 'ask me' or 'check with me' and name no tool or device. Not for clarifying what they just asked, or for anything you can work out yourself. Anyone who has the link can answer. A receipt records what was shown and chosen, not who answered. Use sess_ask for one question with up to 12 simple fields and one button, or to ask many people through one link; use sess_request when the person must see records or choose between buttons such as Approve and Reject. Shortcut for sess_request: give a question and up to 12 simple fields; it creates the same hosted form and returns immediately with a private link (interaction.humanUrl) and a handle, in state "waiting". It never waits and never returns an answer. Continue exactly as after sess_request: give the link only to the intended person, then call sess_response. Do not call it again for the same question. Keep it short: the question is a short heading, about 40 characters (at most 80); the button a verb of 1-3 words (at most 24); labels and options at most 60. A value over its limit is refused with INVALID_ARGUMENTS naming the field (a JSON pointer), the limit, its length and the fix; nothing is truncated. Shorten that value and call again. Advanced: to ask many people through one link, add respondents with max. Share that link with everyone who should answer; each person gets their own page. Read their answers with sess_response. Example: {"question":"Where should we go on Saturday?","details":"I will check the forecast and suggest a plan.","fields":[{"name":"place","label":"Where to?","type":"choice","options":["Beach","Mountains","City"]},{"…
| Name | Type | Req | Description |
|---|---|---|---|
| details | string | – | A paragraph of plain text (never markup): what the person needs to know; at most 4000. |
| expires_in | integer | – | Advanced: how long the link stays open, in whole seconds from creation: from 30 up to the plan’s link lifetime. Leave it out to use the plan’s lifetime. |
| fields | array | yes | – |
| question | string | yes | The page's heading: a question or a task, about 40 characters; at most 80. |
| respondents | object | – | Advanced: one link that many people can answer, each on their own page. Leave it out to ask one person. A respondent is a browser, not a person. Read their answers with sess_response. |
| submit_label | string | – | A button: a verb, 1–3 words; at most 24. Default "Send". |
| Name | Type | Req | Description |
|---|---|---|---|
| interaction | object | yes | – |
{"details":"I will check the forecast and suggest a plan.","fields":[{"label":"Where to?","name":"place","options":["Beach","Mountains","City"],"type":"choice"},{"label":"Anything to avoid?","name":"avoid","required":false,"type":"text"}],"question":"Where should we go on Saturday?","submit_label":"Plan it"} sess_complete Finish the Sess interaction ~111
Required once you are done with the person’s answer: until you call it, their page keeps waiting for your next step. It closes the page to new answers and does not delete what was given. Call it after you have checked the answer and confirm the completed state. If the outcome is unknown, repeat the identical call.
| Name | Type | Req | Description |
|---|---|---|---|
| expected_revision | integer | yes | – |
| handle | string | yes | Opaque handle from sess_request. |
| response_event_id | string | yes | interaction.response.eventId from the verified answer. |
| Name | Type | Req | Description |
|---|---|---|---|
| interaction | object | yes | – |
No examples provided.
sess_publish Show the next Sess step ~159
Replace the form with the next interaction after a verified answer, keeping the same private link. Requires the verified answer for expected_revision. If the outcome is unknown, repeat the identical call.
| Name | Type | Req | Description |
|---|---|---|---|
| expected_revision | integer | yes | – |
| handle | string | yes | Opaque handle from sess_request. |
| interaction | object | yes | A Sess form: a title, components shown in order and 1-5 buttons. Keep it short: a person reads it on a phone. A value over its limit is refused with INVALID_ARGUMENTS naming the field (a JSON pointer… |
| response_event_id | string | yes | interaction.response.eventId from the verified answer. |
| Name | Type | Req | Description |
|---|---|---|---|
| interaction | object | yes | – |
No examples provided.
sess_request Ask a person through Sess ~399
Use it when you need the person's approval, a decision or details only they can give before you go on, even if they only say 'ask me' or 'check with me' and name no tool or device. Not for clarifying what they just asked, or for anything you can work out yourself. Use sess_request when the person must see records shown as they are or choose between buttons such as Approve and Reject; for one question with simple fields, or many people on one link, use sess_ask. Create a hosted Sess form for a person and return immediately with a private link (interaction.humanUrl) and a handle. Give the link only to the intended person in this private conversation; anyone holding it can answer. A receipt records what was shown and chosen, not who answered. Then call sess_response with the handle and revision. Do not call sess_request again for the same question. Keep it short: a title of about 40 characters (at most 80), button labels a verb of 1-3 words (at most 24), field labels at most 60, help one hint (at most 160). A value over its limit is refused with INVALID_ARGUMENTS naming the field (a JSON pointer), the limit, its length and the fix; nothing is truncated. Shorten that value and call again.
| Name | Type | Req | Description |
|---|---|---|---|
| expires_in | integer | – | Advanced: how long the link stays open, in whole seconds from creation: from 30 up to the plan’s link lifetime. Leave it out to use the plan’s lifetime. |
| interaction | object | yes | A Sess form: a title, components shown in order and 1-5 buttons. Keep it short: a person reads it on a phone. A value over its limit is refused with INVALID_ARGUMENTS naming the field (a JSON pointer… |
| Name | Type | Req | Description |
|---|---|---|---|
| interaction | object | yes | – |
No examples provided.
sess_response Read the person’s Sess answer ~335
Wait a bounded time (default and maximum 45 s; wait_ms 0 reads once) for the answer to exactly this revision. state "answered" means the answer passed full Sess receipt verification. state "expired" means the link closed with no answer: it is final; create a new form if you still need one. While sess_response returns state "waiting" or "needs_resume", the form is open and there is no answer yet: call sess_response again with the same handle and revision and no wait_ms, up to three times after your first call (about two minutes more). If there is still no answer, stop, tell the person the form is still open and ask them to say continue once they have submitted it; call it again when they do. Do not poll beyond that, and do not ask the same question again. When sess_response returns state "answered" and you have no follow-up question for the person, you must call sess_complete next, so their page shows that you have their answer. Call sess_publish instead only to show them a follow-up. For a handle from sess_ask with respondents (an Ask), it reads once, without waiting, and returns interaction.ask: whether it is open, the counts and each person’s verified answer so far. Read it again later for more answers; state "closed" means no more answers will come. Do not call sess_publish or sess_complete for an Ask: each person’s page completes on its own.
| Name | Type | Req | Description |
|---|---|---|---|
| handle | string | yes | Opaque handle from sess_request. |
| revision | integer | yes | – |
| wait_ms | integer | – | – |
| Name | Type | Req | Description |
|---|---|---|---|
| interaction | object | yes | – |
No examples provided.
How do I install the Sess MCP server?
Sess runs locally as an npm package, launched with npx -y @sess.si/mcp. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.
npm · @sess.si/mcp
claude mcp add si-sess-sess -- npx -y @sess.si/mcp
What is the Sess MCP server?
Sess is an MCP server listed in the public MCP registry as si.sess/sess. Let your agent ask people for answers and approvals by link, with signed receipts. This page covers its npm package (@sess.si/mcp).
Is the Sess MCP server safe to use?
Sess scores 83 out of 100 on VerifyMCP. We recorded 1 known advisory against it as of 21 November 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 Sess MCP server expose?
Sess exposes 5 tools: sess_request, sess_response, sess_publish, sess_complete, sess_ask. Their descriptions and schemas cost roughly 1,663 tokens of context every time the server is loaded.
Is the Sess MCP server still maintained?
Sess is still listed as active in the MCP registry. We last reached this channel on 21 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 Sess MCP server under?
Sess 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.
What did VerifyMCP test on the Sess MCP server?
VerifyMCP ran 16 checks on Sess, covering Malware scan, Known CVEs, Install scripts, Dependency health and 12 more. Endpoint Security and Transport & Reachability do not apply to an npm package, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.