Datasets:
task_id stringlengths 6 13 | domain stringclasses 8
values | prompt stringlengths 927 2.61k |
|---|---|---|
finance-2 | finance | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Input data: prices.csv contains date,ticker,close columns with unadjusted quotes in each ticker's n... |
industrial-1 | industrial | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The pressure log for the pump skid is in skid_log.csv with columns ts,pressure,unit, typically one ... |
math-1 | math | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Answer every exact-integer problem listed below. Because the values are too large for a naive brute... |
media-8 | media | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The file fps.txt contains the frame rate, using non-drop timecode format HH:MM:SS:FF. The file clip... |
office-1 | office | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Act in the capacity of the accounts-payable clerk. The provided inputs are vendors.csv (containing ... |
science-2 | science | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
seqs.fasta contains DNA records where the id is the first header word; sequences may be lower-case/... |
security-1 | security | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Analyze the sshd logs located in the workspace: auth.log.1 (older half) and auth.log (newer half), ... |
software-2 | software | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
In `ttlcache.py`, build the `TTLCache` class (an LRU cache featuring per-entry time-to-live) strict... |
finance-3 | finance | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Files: prices.csv (date,ticker,close: unadjusted quotes in each ticker's own currency; some rows ar... |
industrial-3 | industrial | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Simulate to tune a discrete PID controller for tank level. Implement the plant yourself using forwa... |
math-3 | math | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Resolve the following computational problems precisely (python3 and numpy are available; Python int... |
media-10 | media | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Construct an RSS 2.0 feed from markdown posts. The file site.json contains the fields title, base_u... |
office-2 | office | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Meeting notes are located in notes/meeting_*.txt, and the team roster is in roster.csv (columns: fu... |
science-3 | science | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Files provided: seqs.fasta contains DNA records where the id is the first word of the header; seque... |
security-6 | security | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Perform defensive log triage on a synthetic web server. The file access.log uses the combined log f... |
software-4 | software | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
`old_report.py` is a disposable script. Remove it and create a formal command-line utility named `l... |
finance-9 | finance | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The file transactions.csv, with columns txn_id,date,symbol,side,qty,price, contains BUY and SELL tr... |
industrial-12 | industrial | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
To minimize the makespan, assign jobs to machines. jobs.csv: job,family,proc,release (minutes). set... |
math-6 | math | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The file roads.csv defines a road network where lines beginning with # are treated as comments, and... |
media-11 | media | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Create an RSS 2.0 feed derived from markdown posts. The site.json file contains the title, base_url... |
office-5 | office | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Distribute meeting requests into rooms. Input files: rooms.csv (room,capacity,projector,open,close)... |
science-11 | science | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
signal.txt contains 3000 samples (one per line) from a sensor sampled at 1000 Hz. The data consists... |
security-7 | security | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
This task involves defensive log triage for a synthetic web server. The file access.log uses the co... |
software-5 | software | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Discard the temporary `old_report.py` script. Create a new command-line utility named `logq.py`, in... |
finance-13 | finance | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Value American options using a Cox-Ross-Rubinstein binomial tree. market.json contains spot, rate (... |
industrial-13 | industrial | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Arrange jobs across machines to minimize the makespan. jobs.csv: job,family,proc,release (minutes).... |
math-7 | math | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The file roads.csv contains a road network where lines beginning with # are comments, and if a pair... |
media-12 | media | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
stream.mcf is a corrupted media container. Its structure consists of the magic bytes 'MCF1' followe... |
office-11 | office | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The file contacts.csv (columns: contact_id,name,email,phone,company,title,updated) contains duplica... |
science-12 | science | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Combine the chemical kinetics model found in params.json (containing rate constants k1,k2,k3) with ... |
security-9 | security | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The task involves building detection rules for synthetic shell and process command lines. The file ... |
software-9 | software | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The workspace contains a package index (`index.json`: package -> version -> list of dependency spec... |
finance-17 | finance | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Perform Brinson-Fachler performance attribution across 4 periods. The input files are as follows: h... |
industrial-17 | industrial | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Calibrate and execute a Kalman filter. Input file obs.csv contains columns t (seconds), sensor ('po... |
math-14 | math | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The alphabet consists of {a,b,c}. The file params.json specifies the queries (lengths n) and the fo... |
media-13 | media | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The file stream.mcf is a corrupted media container. Its structure begins with the magic bytes 'MCF1... |
office-16 | office | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Create a 14-day schedule beginning on Monday 2026-11-02 (weeks run Mon-Sun, corresponding to days 1... |
science-13 | science | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Using the 4-species kinetics defined in params.json (k1,k2,k3,kf,kr, dose, t_dose), begin integrati... |
security-11 | security | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Perform a defensive dependency audit using synthetic data. The file sbom.json contains components w... |
software-10 | software | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Create the file `calc.py` which must contain `class CalcError(Exception)` and `evaluate(expr, env=N... |
math-17 | math | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Facility location problem. The file depots.csv has columns (id,fixed,cap); activating a depot incur... |
science-14 | science | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The file assay.csv contains enzyme initial-rate data with the columns id,S,S_unit,I_uM,rate,rate_un... |
security-12 | security | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Correlate the following three synthetic logs: auth.csv (containing time in ISO-8601 format, host re... |
software-11 | software | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Create a file named `calc.py` that contains the definition of `class CalcError(Exception)` along wi... |
science-16 | science | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The file dist.csv contains a symmetric 13x13 matrix of estimated pairwise evolutionary distances, w... |
security-14 | security | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The file trace.csv contains a synthetic TCP packet capture with columns named ts, src, sport, dst, ... |
software-14 | software | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The queueing model relies on `sim.py`, a compact discrete-event simulator featuring a `Sim` class (... |
security-15 | security | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
trace.csv contains a synthetic TCP packet trace with columns ts, src, sport, dst, dport, seq, flags... |
software-15 | software | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
The file `sim.py` implements a compact discrete-event simulation system, providing a `Sim` class wi... |
software-18 | software | Operate from the current directory by executing shell commands; the environment provides python3 with access to numpy, pandas and Pillow. Once completed, respond with {"answer": "done"} to trigger scoring of your files.
Create a stub for `patchlib.py` containing: `PatchError(Exception)`, `make_patch(old, new, context=... |
Mixed Skills v3: eight-domain verified workspace tasks
One OpenEnv server, 50 deterministic procedurally generated tasks across the eight evaluation domains: software engineering, industrial and physical systems, natural science, office work, finance and economics, math and formal reasoning, cybersecurity, and media production. Each task gives the agent a scratch workspace with input files and a written specification. The agent works with shell commands and is scored by a hidden verifier on the files it leaves behind.
tasks.jsonl lists every task id with its domain and the exact first-turn prompt. The image is
docker.io/chrismun/mixed-env:v3 (linux/amd64); it contains 150 task ids and the submission names the 50 used for training.
Interaction
Each turn the model replies with exactly one JSON object:
{"command": "<shell command>"}runs in the workspace (python3 with numpy, pandas and Pillow). At most 12 commands per episode; each command's output is truncated to its last 1500 characters, so strong solutions inspect withheadand do the work in a script.{"answer": "done"}ends the episode and scores the workspace.
Observations carry the command output, the commands left and. There is no reminder text. Replies that are not a single JSON object end the episode, which is then scored on the workspace as it stands.
Reward
Terminal and in [0, 1]. Every verifier is a weighted set of independent sub-checks recomputed from the task inputs, so partial progress earns partial reward and only a fully correct deliverable scores 1.0. References live only inside the image, agent commands run as an unprivileged user with memory and file-size limits, and verifiers that execute agent code do so in a subprocess with a timeout. An untouched workspace scores at most 0.25 on every task.
Calibration
Tasks were selected by measuring Qwen3.8-27B (thinking off, strict one-JSON-per-turn replies, temperature 0.6 to match the arena) with 8 to 16 rollouts per task. The target was a hard start: a mean starting pass@1 near 0.3, with groups of rollouts that still carry reward variance.
- Final 50 tasks, fresh 16-rollout pass at T=0.6 (800 rollouts, none of them used for selection): mean pass@1 0.29, estimated probability that a group of four rollouts has reward variance 0.68, 5.2 turns per episode. 43 of 50 tasks are fully solved at least once; 6 tasks sit below 0.1 and 2 above 0.7.
- Every task runs bare: no reply-format reminder and no per-step hint, so replies must be strict single JSON objects unprompted.
- Prompt text: the task statements and the three global strings were rewritten by Qwen3.8-27B from the original statements under strict gates (every number, file name, identifier, quoted string and JSON example must survive verbatim). Rewritten difficulty was not re-matched task by task for the newest 28 tasks.
- No transcript in the profiling runs touched verifier or task source.
Domains
| Domain | Skills exercised |
|---|---|
| software | multi-file bug fixing against hidden tests, implementing modules from a spec, CLI conversion, migration scripts, simulator debugging, performance and dependency problems |
| industrial | sensor log analysis and fault reports, control tuning by simulation, pipe and pump sizing, tariff billing, inventory policy, flow-line simulation |
| science | curve and rate fitting, sequence analysis and protein mass, ODE integration, hypothesis tests, titration equilibria, spectral analysis |
| office | payables reconciliation, meeting-minutes extraction, room scheduling, expense audits, payroll, contact deduplication |
| finance | amortization, market-data analytics, ledger audit, bond pricing, tax lots, FX arbitrage |
| math | exact number theory, proofs by computation, assignment optimization, shortest routes, constructive codes, absorbing Markov chains, impartial games |
| security | log incident analysis, hardening vulnerable code, layered-encoding and crypto forensics, detection rules, advisory matching (all synthetic, defensive) |
| media | subtitle retiming, image pipelines, audio synthesis, SVG charts, timecode conformance, markdown to RSS |
Files
tasks.jsonl:task_id,domain,promptfor each of the 50 training tasks, in training order (domains interleaved).
- Downloads last month
- 77