Haiku 5.5
Native agent · Checked operation passedChecked operation <guest-path> · local stdio MCP start + initialize + tools/list
Agent-reported scope: Install mcp-biomodelling-servers from the source checkout and drive the MaBoSS MCP server over stdio: call create_session with a label and read the returned session ID and label.
MCP server returned: MaBoSS · version 2.5.2.
MCP tools returned: edit_maboss_readouts, export_handoff_bundle, create_session, list_sessions, list_artifact_sessions, set_default_session, delete_session, import_neko_handoff, bnet_to_bnd_and_cfg, build_simulation, run_simulation, export_maboss_bnd_cfg, export_maboss_handoff, get_maboss_nodes, get_maboss_initial_state, get_maboss_logical_rules, change_maboss_rule, get_maboss_mutations, update_maboss_parameters, set_maboss_output_nodes, set_maboss_initial_state, simulate_mutation, visualize_network_trajectories, get_simulation_result, list_generated_files, clean_generated_files. Tool listing does not establish authenticated tool-call success.
Run details
- run
- marcorusc--mcp-biomodelling-servers-claude-haiku-5-5-1793176561604258445
- date
- Oct 28, 2026
- time to working
- 383.3s
- product hosting / execution
- Local stdio MCP server in the guest
- recorded operation scope
- product operation
- model
- claude-haiku-5-5
- agent / version
- claude-code / 2.1.310 (Claude Code)
- platform
- linux
- reported cost
- $0.013667
- Recorded source
- https://github.com/marcorusc/mcp-biomodelling-servers
- Backend
- hyperv
- Tool transport
- mcp-assigned-vm
- Requested source pin (not an observed revision)
- {"commit": "d9079ee2eb8965d27939631da98117abe774091b", "url": "https://github.com/marcorusc/mcp-biomodelling-servers"}
- Test version
- readme-2d57868128a84902c09a58991c46c5fe288928e55b1de8282ab5003e13159c2f
- revision
- d9079ee2eb8965d27939631da98117abe774091b
Safety: No problems traced to this repository code in these checks.
No problems traced to this repository code in these checks.
Reviewed as expected for a test run: attempted writes that failed, so nothing was written.
OSV scanned supported lockfiles; unpinned dependencies and unsupported formats are outside this check.
A short run cannot establish that code is safe.
Connections show IP addresses; proxies can hide the final hostname.
Tracing covers dispatched commands and their children, not the whole OS; guest root can interfere with tracing.
Answers recorded for this run
- Each server can maintain multiple isolated modelling sessions; a MaBoSS session is created and its label is returned: Real input produced the expected result: coldpass-claim-5521
Source: Each server can maintain multiple isolated modelling sessions.
Input: {"tool": "create_session", "arguments": {"label": "coldpass-claim-5521"}}
Expected: session created with that label returned
Command: <guest-path> tools/call {"tool": "create_session", "arguments": {"label": "coldpass-claim-5521"}}
Observed: …coldpass-claim-5521) (set as default)", "type": "text"}], "isError": false, "structuredContent": {"result": "Session created: 36fd9abe-31c8-44f2-9447-3f8f4c514711 (coldpass-claim-5521) (set as default)"}}
Local or hosted: local
What the agent did
Install mcp-biomodelling-servers from the source checkout and drive the MaBoSS MCP server over stdio: call create_session with a label and read the returned session ID and label.
cd <guest-path> && pwd && ls -la && git status 2>&1 | head -3; python3 --version; cat README.md 2>/dev/null | head -c 6000
curl -s -m 10 -i $CDP_URL/json/version; echo; echo "exit=$?"; echo "CDP_URL=$CDP_URL"; ls /inputs 2>&1
cd <guest-path> && sed -n '/### Enabling/,/## /p' README.md | head -c 100 >/dev/null; grep -n '^#' README.md; ls .coldpass; ls <guest-path> 2>&1
cd <guest-path> && sed -n '146,220p;258,300p' README.md
python3 -m venv <guest-path> 2>&1 | tail -3; ls <guest-path> 2>&1 | head
cd <guest-path> && timeout 900 <guest-path> -m pip install --upgrade pip ><guest-path> 2>&1; echo rc=$?; timeout 1500 <guest-path> -m pip install "mcp-biomodelling-servers" > <guest-path> 2>&1; echo rc=$?; tail -5 <guest-path>
cd /tmp && for s in neko maboss physicell biomass; do <guest-path> --version; echo "rc=$?"; done; <guest-path> --help | head -30
head -c 3000 <guest-path>; echo; grep -n "def \|argparse" <guest-path> | head -60
cd <guest-path> && ls MaBoSS mcp_biomodelling_servers; grep -rn "@mcp.tool\<email>\|def create\|def list_" MaBoSS/*.py | head -40
cd <guest-path> && sed -n 185,240p MaBoSS/server.py
cd <guest-path> && grep -n "All four servers use MCP over stdio\|Each one serves MCP over stdio\|Each server can maintain multiple\|Prebuilt wheels cover\|No system software is needed\|python -m pip install mcp-biomodelling-servers\|installs four console\|The installation provides four console" README.md
cd <guest-path> && cat > probe.json <<'EOF'
{"kind":"mcp","argv":["<guest-path>"],"cwd":"<guest-path>","tool":"create_session","arguments":{"label":"coldpass-probe-7319"},"expected_text":"coldpass-probe-7319"}
EOF
cat > claims.json <<'EOF'
[{"kind":"claim","claim":"Each server can maintain multiple isolated modelling sessions; a MaBoSS session is created and its label is returned","source":"Each server can maintain multiple isolated modelling sessions.","input":"create_session with label coldpass-claim-5521","expected":"session created with that label returned","probe_kind":"mcp","argv":["<guest-path>"],"tool":"create_session","arguments":{"label":"coldpass-claim-5521"},"expected_text":"coldpass-claim-5521","cwd":"<guest-path>"}]
EOF
python3 -m json.tool probe.json >/dev/null && python3 -m json.tool claims.json >/dev/null && echo ok
cd <guest-path> && <env-var>=<guest-path> timeout 900 python3 -I <guest-path> --validate --replay 2>&1 | tail -c 5000; echo "rc=${PIPESTATUS[0]}"
cd <guest-path> && grep -n -i "pytest\|test\|cron\|--once\|--now\|webhook\|schedul" README.md | head -20; ls tests | head -30; ls .github/workflows 2>/dev/null
cd <guest-path> && cat > cycles.json <<'EOF'
{"detected":false,"checks":[],"reason":"Inspected README.md (Installation, Configure an MCP client, Sessions/artifacts, Run from source) and the four console entry points (mcp-neko-server, mcp-maboss-server, mcp-physicell-server, mcp-biomass-server). They serve MCP over stdio and expose no cron, loop, webhook, --once or --now trigger. The only CI schedule is in .github/workflows, which is not a product cycle."}
EOF
cat > examples.json <<'EOF'
[]
EOF
python3 -m json.tool cycles.json >/dev/null && python3 -m json.tool examples.json >/dev/null && echo ok
cd <guest-path> && <env-var>=<guest-path> timeout 900 python3 -I <guest-path> --validate --replay 2>&1 | head -c 1500; echo; echo "rc=${PIPESTATUS[0]}"
cd <guest-path> && grep -n '^version\|^name' pyproject.toml; timeout 1500 <guest-path> -m pip install --force-reinstall --no-deps "." > <guest-path> 2>&1; echo rc=$?; tail -2 <guest-path>; cd /tmp && <guest-path> --version
cd <guest-path> && ls -d build *.egg-info 2>/dev/null; find . -name "*.egg-info" -maxdepth 2 | head; <env-var>=<guest-path> timeout 900 python3 -I <guest-path> --validate --replay 2>&1 | head -c 400; echo; echo "rc=${PIPESTATUS[0]}"
cd <guest-path> && cat > report.json <<'EOF'
{"outcome":"working","fault_owner":null,"operation":"Install mcp-biomodelling-servers from the source checkout and drive the MaBoSS MCP server over stdio: call create_session with a label and read the returned session ID and label.","operation_scope":"Linux guest, Python 3.12.3 venv at <guest-path> (README requires Python 3.10-3.15). Covers the MaBoSS server only; NeKo, PhysiCell and BioMASS entry points were installed and report version 2.5.2 but were not exercised. No browser needed; CDP endpoint http://<internal-ip>:9226/json/version returned HTTP 200 (Chrome/155.0.8059.84) but no browser operation was run.","steps":["python3 -m venv <guest-path>","<guest-path> -m pip install --upgrade pip","<guest-path> -m pip install mcp-biomodelling-servers (initial install, then replaced)","<guest-path> -m pip install --force-reinstall --no-deps . (run from <guest-path>, README 'Run from source')","<guest-path> --version -> mcp-maboss-server (mcp-biomodelling-servers) 2.5.2","MCP tools/call create_session {\"label\":\"coldpass-probe-7319\"} via <guest-path> -> 'Session created: <uuid> (coldpass-probe-7319) (set as default)'","<env-var>=<guest-path> python3 -I <guest-path> --validate --replay -> valid true, replay PASS"],"deviations":["Preflight path /inputs/coldpass-preflight.py does not exist in this VM; used <guest-path> as instructed by the checker path.","First installed the PyPI release of 2.5.2, then reinstalled from the workspace source as README 'Run from source' describes; both report version 2.5.2 and the final replay used the source build.","Workspace has no .git metadata; the pinned git fetch/reset was not performed because the source was already present and no git operation was needed to run the product. The pinned commit d9079ee2eb8965d27939631da98117abe774091b could not be verified in the guest.","README documents no test command or example invocation beyond --help/--version, so examples.json is empty; the repo's offline pytest suite was not run as product evidence.","Cycles: no cron, loop, webhook, --once or --now behavior in the product, so detected=false."],"stuck_step":null,"main_blocker":null,"plain_failure":null,"suggested_fix":"None required for the exercised MaBoSS path. Optionally add a documented example command to the README so examples can be replayed."}
EOF
python3 -m json.tool report.json >/dev/null && echo ok; ls -la <guest-path>run JSON · checker evidence · agent report + check · transcript excerpt · error excerpt unavailable