A governed-delegation analytics and decision-support toolkit.
Project description
minimal-oversight
A governed-delegation analytics and decision-support toolkit.
Companion package to "Minimal Oversight: A Theory of Principled Autonomy Delegation" (Azevedo, 2026).
Documentation | Paper | Notebooks
What it does
Helps practitioners answer six questions about their multi-agent pipelines:
- Can this pipeline meet my target quality? — feasibility check
- Where should I place review effort? — water-filling allocation
- Which nodes are most dangerous? — topology and masking analysis
- How much autonomy can I safely grant? — autonomy buffer
- When should humans intervene? — intervention scheduling
- What should stop being delegated? — scope recommendations
Quick start
from minimal_oversight import analyze_pipeline
from minimal_oversight.models import Node, PipelineGraph, AggregationType
# Define your pipeline
gen = Node("generator", sigma_skill=0.55, catch_rate=0.65, review_capacity=0.50)
rev = Node("reviewer", sigma_skill=0.55, catch_rate=0.65, review_capacity=0.50)
merge = Node("merge", sigma_skill=0.55, catch_rate=0.65,
aggregation=AggregationType.PRODUCT)
pipeline = PipelineGraph([gen, rev, merge])
pipeline.add_edge("generator", "reviewer")
pipeline.add_edge("reviewer", "merge")
# One call — full analysis
report = analyze_pipeline(pipeline, p_min=0.80)
print(report)
Install
pip install -e ".[dev]"
Package structure
| Module | Purpose |
|---|---|
models |
Node, PipelineGraph, GovernancePolicy, WorkflowTrace |
estimation |
Infer σ_raw, σ_corr, M*, catch rate, drift from logs |
capacity |
C_op, B_eff, feasibility checks, H_crit |
topology |
Motif detection, delegation centrality, fragility |
allocation |
AMO solver, scope selection, governance recommendations |
intervention |
T*_auto, intervention schedule, alerts, failure diagnosis |
viz |
Pipeline risk plots, masking dashboard, buffer view |
simulation |
Synthetic engine for what-if analysis (subordinate) |
What it is not
- Not an agent framework
- Not a workflow orchestrator
- Not just a plotting library
- Not just the paper's reproduction code
It is a governed-delegation analytics and decision-support library.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file minimal_oversight-0.1.0.tar.gz.
File metadata
- Download URL: minimal_oversight-0.1.0.tar.gz
- Upload date:
- Size: 92.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/6.1.0 CPython/3.13.7
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
7c4b886a5c68a14351ba78873a0d1d0f8521c1356c61643d4eeb708cba4b4b08
|
|
| MD5 |
1cf8cb0ed2dafd217163218a5b3fd11a
|
|
| BLAKE2b-256 |
00d7f9f62449f6b5d4223ddf0caad5dfb5157983135f5385c443e98b55e8c72b
|
Provenance
The following attestation bundles were made for minimal_oversight-0.1.0.tar.gz:
Publisher:
publish.yml on crbazevedo/delegation-lab
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
minimal_oversight-0.1.0.tar.gz -
Subject digest:
7c4b886a5c68a14351ba78873a0d1d0f8521c1356c61643d4eeb708cba4b4b08 - Sigstore transparency entry: 1202155208
- Sigstore integration time:
-
Permalink:
crbazevedo/delegation-lab@516a749814470e2e45b3201e93197b6bff6d4c40 -
Branch / Tag:
refs/tags/v0.1.0 - Owner: https://github.com/crbazevedo
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish.yml@516a749814470e2e45b3201e93197b6bff6d4c40 -
Trigger Event:
release
-
Statement type:
File details
Details for the file minimal_oversight-0.1.0-py3-none-any.whl.
File metadata
- Download URL: minimal_oversight-0.1.0-py3-none-any.whl
- Upload date:
- Size: 45.7 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/6.1.0 CPython/3.13.7
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
bbb9dad053d0ced288e149c8027f8bc970f9ac9caaa194503d485571d33d397b
|
|
| MD5 |
91d5563aac14628e75f124325edfb762
|
|
| BLAKE2b-256 |
94ead8f842c828366b4144acbc321750431ef21833aef931b3907ef82a50abf4
|
Provenance
The following attestation bundles were made for minimal_oversight-0.1.0-py3-none-any.whl:
Publisher:
publish.yml on crbazevedo/delegation-lab
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
minimal_oversight-0.1.0-py3-none-any.whl -
Subject digest:
bbb9dad053d0ced288e149c8027f8bc970f9ac9caaa194503d485571d33d397b - Sigstore transparency entry: 1202155214
- Sigstore integration time:
-
Permalink:
crbazevedo/delegation-lab@516a749814470e2e45b3201e93197b6bff6d4c40 -
Branch / Tag:
refs/tags/v0.1.0 - Owner: https://github.com/crbazevedo
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish.yml@516a749814470e2e45b3201e93197b6bff6d4c40 -
Trigger Event:
release
-
Statement type: