ModelsAtlas
Explore
Atlas / MCP servers / rajudandigam / agent-inspect

agent-inspectBLOCK

mcp/rajudandigam/agent-inspect
MCP server by rajudandigam · ai agents · ci · cli · Audited 2026-10-12
⚑ Report / re-scanSource ↗

Local evidence debugger and trajectory-test toolkit for TypeScript AI agents: inspect causal runs, catch wrong tool paths in CI, and share safe offline evidence.

Verdict
BLOCK
Grade
F
Trust score
51 /100
Exposed tools
24 18r · 6w · 0d
Transport
—
License
MIT
Stars
165
01

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

agent-inspect Local-first evidence for TypeScript AI agents

See what your agent did. Catch the wrong path in CI. Keep the evidence local.

AgentInspect turns TypeScript agent runs into readable execution trees, deterministic trajectory checks, and portable Evidence v2—without requiring an account, collector, or default upload.

Capture once. Debug, prevent, and share from the same local trace.

No account · no collector · no default upload · metadata-only by default

Website · Docs · npm · GitHub

Support level: Stable — see SUPPORT-LEVELS.md. Network behavior: NETWORK-BEHAVIOR.md.

npm install agent-inspect

Agent code

Read from source at commit c969ddf41d6eOBSERVED · 2026-10-12
02

Connect

Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.

claude-code
claude mcp add agent-inspect-consumer-ts-jest-node16 --env ELASTIC_API_KEY=${ELASTIC_API_KEY} --env GITHUB_TOKEN=${GITHUB_TOKEN} --env STUDIO_INGEST_TOKEN=${STUDIO_INGEST_TOKEN} --env STUDIO_TEST_PASSWORD=${STUDIO_TEST_PASSWORD} -- npx -y agent-inspect-consumer-ts-jest-node16
claude-desktop
{
  "mcpServers": {
    "agent-inspect-consumer-ts-jest-node16": {
      "command": "npx",
      "args": [
        "-y",
        "agent-inspect-consumer-ts-jest-node16"
      ],
      "env": {
        "ELASTIC_API_KEY": "${ELASTIC_API_KEY}",
        "GITHUB_TOKEN": "${GITHUB_TOKEN}",
        "STUDIO_INGEST_TOKEN": "${STUDIO_INGEST_TOKEN}",
        "STUDIO_TEST_PASSWORD": "${STUDIO_TEST_PASSWORD}"
      }
    }
  }
}
03

Exposed tools (24)

18 read · 6 write · 0 destructive.

ToolRiskDescription
agent-inspectreadLocal evidence debugger and trajectory-test toolkit for TypeScript AI agents — framework-faithful execution trees, TraceFacts, TraceContract, Evidence v2, and read-only MCP without a collector or default upload.
askreadquestion runner
badreadbad
boomreadthrows
create_share_checked_evidencewriteCreate in-memory share-checked evidence (artifact-gated).
create_share_safe_bundlewriteCreate an in-memory share-safe bundle manifest and redacted exports.
create_share_safe_reportwriteCreate a share-profile markdown report for one run.
echoreadecho
find_slowest_pathwriteSummarize the slowest steps in one run.
future_toolreadmissing annotations
get_navan_rewardsreadLookup rewards points
get_slowest_pathwriteSummarize the slowest steps in one run.
list_recent_failuresreadList recent failed runs in the configured trace directory.
list_recent_runsreadList recent local trace runs in the configured trace directory.
list_tracesreadList local trace runs in the configured trace directory.
lookup_itemreadLookup an item
probereadprobe
run_checkswriteRun deterministic run.status check for one run.
searchreadfixture search tool
search_docsreadSearch docs
slowreadslow then ok
submitPaymentreadfixture payment tool
xready
zedreadlast target
04

Trust audit

BLOCKgrade F · trust 51/100 Do not install this without reading the findings. The audit found something that could harm you or your machine.

LayerWhat it checksResult
L0Provenance & inventoryFAIL
L1Static analysis of the codeWARN
L2Instruction surface (what it tells the agent)PASS
L3Class-specific surfacePASS
L4Behavioural (sandbox)SKIPPED

What the source does

Filesystem
declared (1 observation(s))
Network
declared (1 observation(s))
Shell
none-observed
Dependencies
not all pinned
Secrets in source
found

Findings (25)

HIGHInventory / provenance · inv.suspicious_name · CWE-1104
packages/core/src/safety/omitted-payload.ts
omitted-payload.ts
Why it matters. member named after an attack tool
Fix. remove or justify
MEDIUMInformation disclosure · disclose.log_secret · CWE-209, CWE-532
packages/cli/src/studio-cmd.ts:219
console.log(`[AgentInspect studio] HTTP ingest enabled (token required on POST routes)`);
MEDIUMHard-coded secrets · secret.anthropic · CWE-798, CWE-321
packages/core/test/evidence/evidence-contract-binding.test.ts:100
expected: "sk-ant-abcdefghijklmnopqrstuvwxyz012345",
MEDIUMHard-coded secrets · secret.anthropic · CWE-798, CWE-321
packages/core/test/evidence/evidence-contract-binding.test.ts:119
required: ["sk-ant-abcdefghijklmnopqrstuvwxyz012345"],
MEDIUMHard-coded secrets · secret.anthropic · CWE-798, CWE-321
packages/redact/test/index.test.ts:322
anthropicLine: "anthropic key sk-ant-abcdefghijklmnopqrstuvwxyz",
MEDIUMHard-coded secrets · secret.anthropic · CWE-798, CWE-321
packages/redact/test/index.test.ts:333
expect(serialized).not.toContain("sk-ant-abcdefghijklmnopqrstuvwxyz");
MEDIUMHard-coded secrets · secret.generic · CWE-798, CWE-321
examples/recipes/redact-share-safe-file/src/index.ts:8
apiKey: "FIXTURE_TOKEN_REDACT_ME",
MEDIUMHard-coded secrets · secret.generic · CWE-798, CWE-321
examples/starters/broken-agent-debugging/demo-agent.mjs:131
apiKey: "sk-demo-not-a-real-key",
MEDIUMHard-coded secrets · secret.generic · CWE-798, CWE-321
packages/cli/test/bundle.test.ts:112
const secret = "sk-bundleRedactedSecret1234567890";
MEDIUMHard-coded secrets · secret.generic · CWE-798, CWE-321
packages/cli/test/redact-residual-policy.test.ts:114
const secret = "sk-fixtureResidualSecretValue123456789";
MEDIUMHard-coded secrets · secret.generic · CWE-798, CWE-321
packages/cli/test/redact.test.ts:46
apiKey: "sk-fixtureSecretValue123456789",
MEDIUMHard-coded secrets · secret.github · CWE-798, CWE-321
packages/redact/test/index.test.ts:184
["value.githubToken", "ghp_abcdefghijklmnopqrstuvwxyz123456", "local"],
MEDIUMHard-coded secrets · secret.github · CWE-798, CWE-321
packages/redact/test/index.test.ts:323
githubLine: "token ghp_abcdefghijklmnopqrstuvwxyz123456",
MEDIUMHard-coded secrets · secret.github · CWE-798, CWE-321
packages/redact/test/index.test.ts:334
expect(serialized).not.toContain("ghp_abcdefghijklmnopqrstuvwxyz123456");
MEDIUMHard-coded secrets · secret.private_key · CWE-798, CWE-321
packages/redact/test/index.test.ts:188
"-----BEGIN PRIVATE KEY-----\nabc\n-----END PRIVATE KEY-----",
MEDIUMHard-coded secrets · secret.private_key · CWE-798, CWE-321
packages/redact/test/index.test.ts:325
pem: "-----BEGIN PRIVATE KEY-----\nabc\n-----END PRIVATE KEY-----",
LOWInventory / provenance · inv.binary · CWE-1104
docs/assets/showcase/video/check-pass-fail.webm
check-pass-fail.webm
Why it matters. a compiled or binary member cannot be reviewed from source
Fix. ship source, or explain the binary in the README
LOWInventory / provenance · inv.binary · CWE-1104
docs/assets/showcase/video/debug-tree.webm
debug-tree.webm
Why it matters. a compiled or binary member cannot be reviewed from source
Fix. ship source, or explain the binary in the README
LOWInventory / provenance · inv.binary · CWE-1104
docs/assets/showcase/video/evidence-bundle.webm
evidence-bundle.webm
Why it matters. a compiled or binary member cannot be reviewed from source
Fix. ship source, or explain the binary in the README
LOWInventory / provenance · inv.binary · CWE-1104
docs/assets/showcase/video/explain.webm
explain.webm
Why it matters. a compiled or binary member cannot be reviewed from source
Fix. ship source, or explain the binary in the README
LOWInventory / provenance · inv.binary · CWE-1104
docs/assets/showcase/video/redact.webm
redact.webm
Why it matters. a compiled or binary member cannot be reviewed from source
Fix. ship source, or explain the binary in the README
LOWInventory / provenance · inv.suspicious_name · CWE-1104
packages/core/test/safety/omitted-payload.test.ts
omitted-payload.test.ts
Why it matters. member named after an attack tool
Fix. remove or justify
LOWFilesystem / path · fs.credential_store · CWE-22, CWE-59
packages/cli/test/safety.test.ts:196
pathHint: "/Users/synthetic/.ssh/id_rsa",
Why it matters. touches a credential store
LOWFilesystem / path · fs.traversal · CWE-22, CWE-59
apps/website/lib/load-doc.ts:33
path.resolve(here, "../../.."),
LOWFilesystem / path · fs.traversal · CWE-22, CWE-59
apps/website/lib/load-doc.ts:34
path.resolve(here, "../../../.."),

Gates applied: no_behavioural_pass.

Audited 2026-10-12 · audit v0.4.1 · source sha c969ddf41d6efull audit observations/trust-audit/mcp-server/rajudandigam__agent-inspect.json · Report an issue / request a re-scan
05

Audit history

Every audit this server has had. A grade with a past is a grade somebody is still checking.

DateSourceVerdictGradeScoreChange
2026-10-12c969ddf41d6eBLOCKF51first audit
06

Questions

What is the agent-inspect MCP server?

Local evidence debugger and trajectory-test toolkit for TypeScript AI agents: inspect causal runs, catch wrong tool paths in CI, and share safe offline evidence.

What tools does agent-inspect expose?

24 in total: 18 read-only, 6 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.

Is agent-inspect safe to connect to an agent?

No — not without reading the findings first. The audit graded it F (51/100) and found 1 critical or high issue in the source. Each one is listed on this page with the file and line it is on.

What credentials does agent-inspect need?

It reads ELASTIC_API_KEY, GITHUB_TOKEN, STUDIO_INGEST_TOKEN and STUDIO_TEST_PASSWORD from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.

How current is this page?

The grade is for one exact copy of the source (c969ddf41d6e), read on 2026-10-12. The repository is watched and re-audited when it changes.

Advertisement