MCP server · trust report

io.github.sandeepbazar/ocm-mcp-server

1 check that ran failed — credential hygiene — so this release can’t be trusted as-is.

Publisher
io.github.sandeepbazar
Repository
github.com/sandeepbazar/ocm-mcp-server
Install
pypi:ocm-mcp-server, pypi:ocm-mcp-server, pypi:ocm-mcp-server, oci:ghcr.io/sandeepbazar/ocm-mcp-server:0.4.0
Versions
3
Inspected
4 of 6 checks · 75%
Scored
2026-09-10 · rubric 1.0.0

What we checked

Six static checks, weighted by risk. Every result reflects only what could be observed in the published package and repository — never intent.

6 checks2 clean2 flagged2 not inspected
  • Injection surface

    pass25% of grade

    Tool descriptions/manifest scanned for instruction-injection patterns (imperatives at the model, hidden text, 'ignore previous', data-exfil URLs).

    Scanned 135 text field(s); no instruction-injection patterns found. PASS = no static red flag. A server can serve different tool descriptions at runtime (the rug-pull class), which static scanning cannot see.

  • Supply chain

    pass25% of grade

    Package provenance: namespace verification, repo linkage, maintainer count, account age, postinstall scripts, typosquat distance.

    repository linkage verified

  • Credential hygiene

    fail15% of grade

    How the server takes secrets (env vs plaintext config vs hardcoded); secrets appearing in tool schemas.

    tool "apply_manifestwork" has a secret-shaped input parameter "approval_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "rollback_manifestwork" has a secret-shaped input parameter "approval_token" — secrets must never be tool parameters (the model would supply/handle the credential). tool "apply_cluster_action" has a secret-shaped input parameter "approval_token" — secrets must never be tool parameters (the model would supply/handle the credential).

  • Permission scope

    unscannable15% of grade

    Declared tools vs. breadth (filesystem, network, exec); flags shell-exec and unbounded filesystem access.

    No source files, package.json, or pyproject were fetched; cannot infer capability scope.

  • Version behavior

    warn10% of grade

    Diff of tool definitions between versions; new permissions or changed descriptions in a patch release (the postmark-mcp class).

    vs prior 0.3.0 (minor bump): new tool "list_clusters" appeared; new tool "get_cluster" appeared; new tool "list_cluster_sets" appeared; new tool "list_cluster_set_bindings" appeared; new tool "list_cluster_claims" appeared; new tool "get_cluster_info" touches a sensitive surface (email/exec/net/fs); new tool "get_cluster_health" touches a sensitive surface (email/exec/net/fs); new tool "get_fleet_health" touches a sensitive surface (email/exec/net/fs); new tool "query_events" appeared; new tool "get_pod_logs" appeared; new tool "list_placements" appeared; new tool "get_placement_decision" appeared; new tool "list_addon_placement_scores" appeared; new tool "list_manifestworks" appeared; new tool "get_manifestwork" touches a sensitive surface (email/exec/net/fs); new tool "list_manifestworkreplicasets" appeared; new tool "propose_manifestwork" touches a sensitive surface (email/exec/net/fs); new tool "apply_manifestwork" appeared; new tool "propose_rollback" touches a sensitive surface (email/exec/net/fs); new tool "rollback_manifestwork" touches a sensitive surface (email/exec/net/fs); new tool "list_cluster_management_addons" appeared; new tool "get_addon_health" appeared; new tool "list_addons_for_cluster" appeared; new tool "list_pending_csrs" appeared; new tool "propose_cluster_action" touches a sensitive surface (email/exec/net/fs); new tool "apply_cluster_action" appeared; new tool "list_policies" appeared; new tool "list_policy_violations" appeared; new tool "list_hosted_clusters" appeared; new tool "get_hosted_cluster" appeared; new tool "list_node_pools" appeared; new tool "list_resources" touches a sensitive surface (email/exec/net/fs); new tool "get_resource" touches a sensitive surface (email/exec/net/fs); new tool "list_pending_proposals" appeared; new tool "get_audit_trail" touches a sensitive surface (email/exec/net/fs).

  • Transport config

    unscannable10% of grade

    Remote servers: TLS and auth mode (none/token/OAuth). Local servers: whether the manifest indicates it phones home.

    local stdio server: no remote transport or auth surface, and no phone-home/telemetry signal found in source. Absence of transport findings is NOT a safety guarantee — phone-home hosts can be constructed at runtime and defeat static analysis.

Version history

Each release plotted by grade against the safe line at B. A version that sinks below the line has lost its trusted standing — the shape of a rug-pull.

VersionPublishedGradeScoreInspectedChange
v0.4.0 · current2026-08-13 B804/6 · 75%±0
v0.3.02026-07-29 B1003/6 · 55%±0
v0.2.22026-07-27 B92.53/6 · 55%