Skills/npm/mcp-server

mcp-server@1.2.15

VERIFIED

Publisher claims: "Heroku Platform MCP Server" Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry npm (MCP server)Publisher heroku · publishing for 591 daysInstalls 35,954Last detonated 20 min ago
$claude mcp add mcp-server -- npx -y @heroku/mcp-server
WatchReport an issue

Coverage

Server started and answered tools/list.
Tested
MCP initialize and tools/list · every tool description, for hidden instructions · every file read, file written and host contacted by what ran
Not tested
what the tools do when called with real credentials

Before you give it your token

It will ask for
HEROKU_API_KEYConfiguration, not credentials: HEROKU_HEADERS, HEROKU_MCP_MODE, HEROKU_MCP_SERVER_VERSION, MCP_SERVER_IDLE_TIMEOUT, MCP_SERVER_REQUEST_TIMEOUT
Where it talks to
heroku.com, herokuapp.com (its own)Also test.com, example.invalid — not the publisher's domain; check what for before connecting.
What we saw when it ran
no credential files read · no network
What we could not test
what the tools do when called with real credentials

Connect it with a scoped credential, under the guard, and read the third-party hosts first. The code names test.com, example.invalid as well as its own vendor. The guard injects HEROKU_API_KEY only into requests to heroku.com (and its subdomains) and refuses the rest.

$vet-skills run --secrets HEROKU_API_KEY=heroku.com -- npx -y @heroku/mcp-server

Never paste a credential into a chat, a ticket or a report, including ours. How the guard holds it.

What changed since 1.2.14

vet-skills diff mcp-server 1.2.14 1.2.15

In the sandbox, this version reached for nothing 1.2.14 did not. Compared from the two recorded runs, not from source.

Only one of the two runs produced a tool list; tools are not compared.

The guard can hold an update like this until someone approves it (runtime.on_change: hold). A change that only shows with a real credential, or on a path the run did not reach, is not in a trace and so not in this comparison.

Is [email protected] safe to install?

KillerSkills ran it in the KillerSkills honeypot sandbox (2 scenarios) on 2026-11-20 and scored it 87/100 (Verified). It did what it claims and nothing else we could detect: no undeclared network egress, no credential reads, no persistent instruction writes. Summary: no network, touches nothing outside the project, writes no memory.

Behavior trace

Honeypot run · 37 s · 2 scenarios · 0 deviations
t+CategoryObserved action
0.0sprocessno install scripts in bundle
16sprocessdependency install exit 0
36sprocessexec /app/apps/web/node_modules/.bin/heroku
36sprocessexec /root/.cache/node/corepack/v1/pnpm/10.28.0/dist/node-gyp-bin/heroku
36sprocessexec /app/node_modules/.bin/heroku
36sprocessexec /pnpm/heroku
36sprocessexec /usr/local/sbin/heroku
36sprocessexec /usr/local/bin/heroku
36sprocessexec /usr/sbin/heroku
36sprocessexec /usr/bin/heroku
36sprocessexec /sbin/heroku
36sprocessexec /bin/heroku
36sclaimed33 tools: list_apps, get_app_info, create_app, rename_app, maintenance_on, maintenance_off, get_app_logs, list_private_spaces, list_teams, list_addons, get_addon_info, create_addon, list_addon_services, list_addon_plans, pg_psql, pg_info, pg_ps, pg_locks, pg_outliers, pg_credentials, pg_kill, pg_maintenance, pg_backups, pg_upgrade, ps_list
36stoolsdf964e192e8996e2
36stoolseb25c26aef6f7ddb
36stools081a6d7e441dbadc
36stools784db3fe44d5fa68
36stools33a22610553a37af
36stools0606db21cf37f2f0
36stoolsca7cb6b14b802155
36stools36cd866cf6bf11d0
36stools2149d5fc785dfbd3
36stoolsbec8b9c0127bc0b4
36stools2baacf6658165d08
36stoolsa6f651700554598b
36stoolsd4042d035e94cb85
36stools8c725636a545c7ba
36stools9e1fa10501354df2
36stools9148289469c43302
36stools2c7d7990c907f1c9
36stools815a93d36ef7d7f0
36stools4ebea152215e5fd3
36stools48eb5d0426f802a9
36stoolse11097f617f21d56
36stools44a0a31b8e8e2ecd
36stools6e9bafee6901fde2
36stools22b58c83dd705c44
36stoolsce0e2fafe192e264
36stoolse4194911c7c5d333
36stools9a9817e95e40c715
36stoolsb6c715a3be67b1c9
36stoolseaa8a989cb75de32
36stools65f8e8a58ac1e102
36stoolsdae1366a107d9134
36stools774d7799a1767a97
36stools6966763a42a0be5f
36sprocessscript backend runs setup only
36sprocessno install scripts to re-run

Static findings

MEDIUM
Process kill commanddist/repl/heroku-cli-repl.spec.js:17 · Process kill command — agent can terminate processeskill = sinon.stub();Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
npx invocation of non-well-known packagedist/repl/heroku-cli-repl.spec.js:475 · npx invocation of non-well-known package — may execute untrusted codeit('should fall back to npx when heroku CLI is not available', () => {Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Process kill commanddist/repl/heroku-cli-repl.js:400 · Process kill command — agent can terminate processes// Keep an error sink so a post-kill 'error' doesn't become an uncaught throw.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
npx invocation of non-well-known packagedist/repl/heroku-cli-repl.js:134 · npx invocation of non-well-known package — may execute untrusted code* Falls back to `npx -y heroku@latest repl` only when no suitable local CLI is foundFix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
npx invocation of non-well-known packagedist/repl/heroku-cli-repl.spec.js:447 · npx invocation of non-well-known package — may execute untrusted code// npx is also available, but must not be used when a suitable local CLI existsFix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Process kill commanddist/repl/heroku-cli-repl.spec.js:200 · Process kill command — agent can terminate processes// kill a process that now has work to do.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
npx invocation of non-well-known packagedist/repl/heroku-cli-repl.spec.js:495 · npx invocation of non-well-known package — may execute untrusted codeit('should fall back to npx when the local heroku CLI is too old (< 10.10.0)', () => {Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
npx invocation of non-well-known packagedist/repl/heroku-cli-repl.js:136 · npx invocation of non-well-known package — may execute untrusted code* null only when neither a usable local CLI nor npx is available.Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
npx invocation of non-well-known packagedist/repl/heroku-cli-repl.js:176 · npx invocation of non-well-known package — may execute untrusted codethis.handleFatal('Startup error: No suitable Heroku CLI found. Install the Heroku CLI (version 10.10.0 or higher) in your PATH, or make npx available. Please install one of them.');Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
npx invocation of non-well-known packagedist/repl/heroku-cli-repl.spec.js:458 · npx invocation of non-well-known package — may execute untrusted code// npx must not even be probed when a suitable local CLI is foundFix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Process kill commanddist/tools/data.js:230 · Process kill command — agent can terminate processes* Registers the pg:kill tool with the MCP server.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
npx invocation of non-well-known packagedist/repl/heroku-cli-repl.spec.js:482 · npx invocation of non-well-known package — may execute untrusted code// npx availableFix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
npx invocation of non-well-known packagedist/repl/heroku-cli-repl.js:128 · npx invocation of non-well-known package — may execute untrusted code* - it avoids the npx cold-start penalty (npm registry resolution, often 5-15s) on every spawn,Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
LOW
Context shift attemptREADME.md:325 · Context shift attempt — trying to change operational mode · informational: not an instruction file- `maintenance_on` - Enable maintenance mode for an app.Fix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
MEDIUM
File permission modification commandpackage.json:86 · File permission modification command — agent changes file ownership/permissions"exec-perms": "shx chmod +x dist/*.js",Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
MEDIUM
Access to sensitive file pathsdist/tools/deploy-to-heroku.spec.js:27 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filesexpect(isSafeSourceRelativePath('/etc/passwd')).to.be.false;Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
File type combined with upload actiondist/tools/data.spec.js:42 · File type combined with upload action — agent may exfiltrate local data files…COMMAND_MAP.PG_PSQL} --app='myapp' --command='SELECT * FROM users' --file='query.sql' --credential='mycred' -- 'HEROKU_POSTGRESQL_RED'`); }); it('should handle error response', async () =…Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07

Cost to the agent

Instructions loaded
≈44 tok
1 file · 54 in bundle
Tool list per prompt
≈5,053 tok
33 tools · 153 per tool
Handshake
810 ms
launch to tools/list, stdio
Undocumented tools
0
description under 20 chars · avg 88

Tokens estimated at four characters each. The tool list is re-read by the agent on every turn, so it is the number to compare between servers that do the same job: ranked across the index.

Known vulnerabilities

Not checked against OSV yet; the weekly check has not reached this version.