agent-learning
Evidence-driven decision SDK for AI agents. Each recurring agent-task decision has one small, interpretable TaskPolicy over explicit executable alternatives.
TaskPolicies model reusable decisions among executable alternatives such as models, skills, tools, workflows, or workloads. Factual questions, ordinary chat, reporting, and learning automation are not policy tasks.
How it works
The SDK improves decisions without LLM weight fine-tuning. There are no GPU fine-tune jobs and no opaque update cycles. Four pieces run in the existing Python process:
-
TaskPolicy owns
Ndiscrete actions and one persisted decision authority.lowselects from learned softmax evidence;fullevaluates a structured DecisionFrame against the same action set. -
DecisionResolver applies hard constraints, confidence-weighted Bayesian evidence aggregation, Pareto elimination, robust utility, and information needs. A close result requires an explicit accept/reject tie-break.
-
Score evaluates each episode on-device for intent resolution, task adherence, and task completion. Azure AI evaluators remain opt-in.
-
Learner applies REINFORCE-with-baseline to low-authority TaskPolicy logits from observed outcomes and accept/reject feedback. Full-authority episodes are scored and audited but are not treated as softmax samples.
task-policy-decide closes the loop at execution time. It returns learned
feedback for low authority or an auditable decision certificate, information
needs, and any required tie-break for full authority. Both routes preserve the
same policy ID, version lineage, and action taxonomy.
It also returns a complexity-proportional autonomy assessment. A persisted profile covers intent ambiguity, context variability, outcome observability, decision impact, reversibility, and mandatory approval; action-space size is derived. The resulting low, standard, high, or critical tier scales required outcomes, Wilson confidence, reward, probability, margin, stable snapshots, and drift-audit rate. Autonomous executions continue learning from observable outcomes, while tier-scaled samples request user feedback to detect drift. An explicit accepted-feedback episode is a separate durable authorization path: it pins that action for the task policy and suppresses future feedback prompts until the user explicitly rejects it.
Every episode, reward, run, and deployment is captured by the configured store — in-memory or local files by default, or Azure Cosmos DB — giving you a complete lineage and audit trail of how the policy evolved over time.
1 maintainer
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 agent_learning-0.8.0.tar.gz.
File metadata
- Download URL: agent_learning-0.8.0.tar.gz
- Upload date:
- Size: 124.8 kB
- Tags: Source
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/7.0.0 CPython/3.13.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
f2875bf2c997be4709f5fe3f32c169bcc5563e4cbe82bb1d253c762fab07f46b
|
|
| MD5 |
dbfb5f501a0f02ed2691a0966b18a822
|
|
| BLAKE2b-256 |
6fa691736506de2605bb914a20a32606b1f9f33ef22e4f6441b0dd8505ea4a50
|
Provenance
The following attestation bundles were made for agent_learning-0.8.0.tar.gz:
Publisher:
publish.yaml on microsoft/agent-learning
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
agent_learning-0.8.0.tar.gz -
Subject digest:
f2875bf2c997be4709f5fe3f32c169bcc5563e4cbe82bb1d253c762fab07f46b - Sigstore transparency entry: 2408761177
- Sigstore integration time:
-
Permalink:
microsoft/agent-learning@5d8e2e87503935fcbc76adf17d6dbd2465c6e741 -
Branch / Tag:
refs/heads/main - Owner: https://github.com/microsoft
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish.yaml@5d8e2e87503935fcbc76adf17d6dbd2465c6e741 -
Trigger Event:
push
-
Statement type:
File details
Details for the file agent_learning-0.8.0-py3-none-any.whl.
File metadata
- Download URL: agent_learning-0.8.0-py3-none-any.whl
- Upload date:
- Size: 119.5 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? Yes
- Uploaded via: twine/7.0.0 CPython/3.13.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
27b01025917ca3e0aa9059ae82d693db5bd7fdbaf0da465a7e93263d686f79a3
|
|
| MD5 |
06e17b182d26a5a59ada0d7245fd31ae
|
|
| BLAKE2b-256 |
ad4fe3c67254ead53ecb87e7d5a23faed968c5872ef8c511ccd9d2a7ecc37fd8
|
Provenance
The following attestation bundles were made for agent_learning-0.8.0-py3-none-any.whl:
Publisher:
publish.yaml on microsoft/agent-learning
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
agent_learning-0.8.0-py3-none-any.whl -
Subject digest:
27b01025917ca3e0aa9059ae82d693db5bd7fdbaf0da465a7e93263d686f79a3 - Sigstore transparency entry: 2408761237
- Sigstore integration time:
-
Permalink:
microsoft/agent-learning@5d8e2e87503935fcbc76adf17d6dbd2465c6e741 -
Branch / Tag:
refs/heads/main - Owner: https://github.com/microsoft
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish.yaml@5d8e2e87503935fcbc76adf17d6dbd2465c6e741 -
Trigger Event:
push
-
Statement type:
