Scientific Agent Skills: A Library of Procedural Knowledge for Research Agents
Paper • 2609.00065 • Published • 1
Dataset Viewer
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
paper_id: int64
artifact_kind: string
source_evidence_sha256: string
inference_note: string
evidence: struct<paper_id: int64, comparison: list<item: struct<arm: string, corpus: string, lens: string, too (... 2330 chars omitted)
child 0, paper_id: int64
child 1, comparison: list<item: struct<arm: string, corpus: string, lens: string, tool: string, records: int64, populatio (... 821 chars omitted)
child 0, item: struct<arm: string, corpus: string, lens: string, tool: string, records: int64, population: struct<b (... 809 chars omitted)
child 0, arm: string
child 1, corpus: string
child 2, lens: string
child 3, tool: string
child 4, records: int64
child 5, population: struct<benign: int64, malicious: int64, contextually_risky: int64, obviously_malicious: int64, unlab (... 13 chars omitted)
child 0, benign: int64
child 1, malicious: int64
child 2, contextually_risky: int64
child 3, obviously_malicious: int64
child 4, unlabelled: int64
child 6, reported: struct<recall: string, precision: string, f1: string, fpr: string, flag_rate: string>
child 0, recall: string
child 1, precision: string
child 2, f1: string
child 3, fpr: string
child 4, flag_rate: string
child 7, verification: string
child 8, positive_count_reconstruction: struct<candidates: lis
...
ld 9, relative_to_spector_added_tp_range: list<item: int64>
child 0, item: int64
child 10, relative_to_spector_added_fp_range: list<item: int64>
child 0, item: int64
child 11, assumption: string
child 12, relation: string
child 13, verification: string
child 5, reported_paired_intervals: list<item: struct<arm: string, corpus: string, text: string, verification: string>>
child 0, item: struct<arm: string, corpus: string, text: string, verification: string>
child 0, arm: string
child 1, corpus: string
child 2, text: string
child 3, verification: string
child 6, methods: struct<unit: string, primary_lens: string, secondary_lenses: list<item: string>, nvidia_version: str (... 124 chars omitted)
child 0, unit: string
child 1, primary_lens: string
child 2, secondary_lenses: list<item: string>
child 0, item: string
child 3, nvidia_version: string
child 4, skill_scanner_revision: string
child 5, shared_backends: list<item: string>
child 0, item: string
child 6, restricted_capabilities: string
child 7, severity: string
child 7, limitations: list<item: string>
child 0, item: string
algorithm: string
files: list<item: struct<file: string, sha256: string, bytes: int64>>
child 0, item: struct<file: string, sha256: string, bytes: int64>
child 0, file: string
child 1, sha256: string
child 2, bytes: int64
to
{'algorithm': Value('string'), 'files': List({'file': Value('string'), 'sha256': Value('string'), 'bytes': Value('int64')})}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
return get_rows(
dataset=dataset,
...<4 lines>...
column_names=column_names,
)
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
File "/src/services/worker/src/worker/utils.py", line 127, in get_rows
rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
File "/src/services/worker/src/worker/utils.py", line 483, in safe_iter
yield from ds.decode(False) if ds.features else ds
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2951, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2461, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2486, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 547, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 430, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
paper_id: int64
artifact_kind: string
source_evidence_sha256: string
inference_note: string
evidence: struct<paper_id: int64, comparison: list<item: struct<arm: string, corpus: string, lens: string, too (... 2330 chars omitted)
child 0, paper_id: int64
child 1, comparison: list<item: struct<arm: string, corpus: string, lens: string, tool: string, records: int64, populatio (... 821 chars omitted)
child 0, item: struct<arm: string, corpus: string, lens: string, tool: string, records: int64, population: struct<b (... 809 chars omitted)
child 0, arm: string
child 1, corpus: string
child 2, lens: string
child 3, tool: string
child 4, records: int64
child 5, population: struct<benign: int64, malicious: int64, contextually_risky: int64, obviously_malicious: int64, unlab (... 13 chars omitted)
child 0, benign: int64
child 1, malicious: int64
child 2, contextually_risky: int64
child 3, obviously_malicious: int64
child 4, unlabelled: int64
child 6, reported: struct<recall: string, precision: string, f1: string, fpr: string, flag_rate: string>
child 0, recall: string
child 1, precision: string
child 2, f1: string
child 3, fpr: string
child 4, flag_rate: string
child 7, verification: string
child 8, positive_count_reconstruction: struct<candidates: lis
...
ld 9, relative_to_spector_added_tp_range: list<item: int64>
child 0, item: int64
child 10, relative_to_spector_added_fp_range: list<item: int64>
child 0, item: int64
child 11, assumption: string
child 12, relation: string
child 13, verification: string
child 5, reported_paired_intervals: list<item: struct<arm: string, corpus: string, text: string, verification: string>>
child 0, item: struct<arm: string, corpus: string, text: string, verification: string>
child 0, arm: string
child 1, corpus: string
child 2, text: string
child 3, verification: string
child 6, methods: struct<unit: string, primary_lens: string, secondary_lenses: list<item: string>, nvidia_version: str (... 124 chars omitted)
child 0, unit: string
child 1, primary_lens: string
child 2, secondary_lenses: list<item: string>
child 0, item: string
child 3, nvidia_version: string
child 4, skill_scanner_revision: string
child 5, shared_backends: list<item: string>
child 0, item: string
child 6, restricted_capabilities: string
child 7, severity: string
child 7, limitations: list<item: string>
child 0, item: string
algorithm: string
files: list<item: struct<file: string, sha256: string, bytes: int64>>
child 0, item: struct<file: string, sha256: string, bytes: int64>
child 0, file: string
child 1, sha256: string
child 2, bytes: int64
to
{'algorithm': Value('string'), 'files': List({'file': Value('string'), 'sha256': Value('string'), 'bytes': Value('int64')})}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Agent Skill Security Research Artifacts
This collection releases derived aggregate evidence, original figure data and plots, offline analysis code, and reproducibility manifests. The current research portfolio has 4 consolidated empirical manuscript directions: scanner configuration/gate comparisons, small-model score interfaces and source transfer, runtime cascade action composition, and deterministic-rule maintenance/evidence contracts. Current authoring/delivery state: Four consolidated empirical manuscripts and standalone supporting supplements; delivery verification is recorded in separate source-bound local receipts. This card does not certify acceptance, deployment validity, or completion of manuscript production.
paper01/reproduction: Decision Policies Change Agent-Skill Scanner Rankings: A Shared-Backend Empirical Audit. How do shared-backend scanner configurations and severity/native/any-finding gates change detection, benign flags and completion accounting?paper03/reproduction: Detection, Workload, and Intervention: Auditing Deterministic Rule Maintenance for Agent Security. Which detection obligations and intervention contracts change under rule maintenance, parser fallback and profile selection?
The paperXX paths also preserve thirteen historical analysis partitions. These are stable evidence identifiers, not thirteen active papers or independent replications. In particular, the current paper03/reproduction companion consolidates static and runtime rule evidence; earlier paper13 remains historical runtime-rule evidence. Small-model and runtime-cascade analyses share two evaluation populations, and repeated revisions or model outputs are not independent security trials. Each central result has one manuscript owner.
Archived measurements, fresh aggregate arithmetic, and exploratory reanalysis are labeled separately. Source-defined positive and block tags are not independently adjudicated authorization truth. A detection flag, advisory confirm, emitted block decision, and actually prevented unauthorized effect are different outcomes. Refitted score cutoffs are descriptive; frozen transfer is a separate analysis. Existing multi-tier guards and deterministic/semantic policies are relevant prior art, so the portfolio claims empirical observations rather than invention of these broad designs.
paper_catalog.csv records each module's relationship to the current portfolio. Protected source payloads, prompts, credentials, original private repository identities, individual score vectors, and manuscript prose are withheld. Version digests bind bytes; they do not validate scientific labels. AI-assisted owner-side working assessments are not external human peer review or editorial decisions.
Reproduction
Use the module-specific reproduction instructions with environment-requirements.txt. Aggregate arithmetic and figure replay require no provider calls or protected inputs. Full row-level or inference replay can require authorized upstream inputs and historically matched harness/tokenizer/configuration dependencies. Recovered current source does not by itself establish the exact historical inference implementation. Public release replay does not imply new inference, training, live tool execution, or human adjudication.
recompute_aggregate_metrics.py checks released confusion-matrix arithmetic; recomputed-confusion-metrics.csv records results. Nominal independent-binomial intervals, where present in legacy modules, are illustrative diagnostics and do not replace dependence-aware or selection-aware analysis. Use the current companion's stated estimands and assumptions. SHA256SUMS.json inventories released bytes. Preserve negative findings and legacy evidence even when their standalone manuscript is retired.
The shared research/analyses/skill directory contains pinned pure-method extraction and inert-fixture replay instructions.
Rights and workflow tools
Derived numerical data and original figures use CC BY 4.0. Original aggregate analysis code is released under the MIT terms in LICENSE-CODE.txt. Third-party sources retain their own terms; this release grants no rights to withheld upstream inputs.
Installed academic skills support writing, citation verification, visualization, statistical analysis, hypothesis development, evidence critique, review and venue formatting. Installed revisions: scientific-agent-skills 154988403bb5a18e9d3c0ce4e6d5e2e4b184a298 and LitReviewSkill a53cd419352e4dd05958f67340fde3642d84abc3. Attribution: Kassis et al., Scientific Agent Skills (2026), workflow repository, and LitReviewSkill. Skills do not certify validity or novelty.
- Downloads last month
- 317