neosloc archive

Repositories /

pydantic/pydantic

https://github.com/pydantic/pydantic · commit 3eca5695dabc · evaluated 2026-10-12 with neosloc 0.8.0 · JSON

1.62 integrability index over 8 applicable dimensions
library surfaces · owns data
$2.7M → $243k classic COCOMO → neosloc value (9% survives)
buy make / buy = 289.27 over 3 years

Dimensions

Interface surface 4 exemplary An API reference is generated from the code (autodoc, mkdocstrings, typedoc, …).

Can another program use it through a documented, machine-readable surface?

library ladder · level 4

Contract stability 3 solid A deprecation mechanism is used in code (warnings, @deprecated, Deprecation/Sunset headers) or the API is versioned.

Can integrators rely on it not changing under them?

  • next: L4 Across tagged releases, no OpenAPI operation, CLI option or __all__ symbol was removed without first being deprecated. (removed without prior deprecation: symbol AmqpDsn (removed in v2.13.0b1), symbol AnyHttpUrl (removed in v2.13.0b1), symbol CockroachDsn (removed in v2.13.0b1))

all ladder · level 3

  • ✓ L1 Releases are versioned (semver tags or versioned changelog sections). 196 semver tags, latest v2.14.1
  • ✓ L2 A changelog exists and marks breaking changes. HISTORY.md marks breaking changes
  • ✓ L3 A deprecation mechanism is used in code (warnings, @deprecated, Deprecation/Sunset headers) or the API is versioned. pydantic-core/python/pydantic_core/core_schema.py 25 matchespydantic/main.py 20 matches
  • ✗ L4 Across tagged releases, no OpenAPI operation, CLI option or __all__ symbol was removed without first being deprecated. removed without prior deprecation: symbol AmqpDsn (removed in v2.13.0b1), symbol AnyHttpUrl (removed in v2.13.0b1), symbol CockroachDsn (removed in v2.13.0b1)
Events n/a Not applicable: no service and no long-running CLI mode (batch CLIs, libraries and desktop apps have no events).

Can others react to state changes without polling?

Identity n/a Not applicable: uses no credentials and has no service surface.

Can a machine act with its own revocable, least-privilege identity?

Data portability 0 absent Not yet: The data schema is explicit (migrations, DDL, schema files).

Can its data get in and out, in bulk, in open formats, with an explicit schema?

  • next: L1 The data schema is explicit (migrations, DDL, schema files). (no migrations or schema files)

all ladder · level 0

  • ✗ L1 The data schema is explicit (migrations, DDL, schema files). no migrations or schema files
  • ✗ L2 An export exists (endpoint or command, not only the UI). no export
  • ✗ L3 A bulk export in an open format, and a bulk import. no open export format
  • ✗ L4 The export format is documented and the round trip is tested. the export format isn't documented
Agent ergonomics 0 absent Not yet: Errors are specific exception types, and library code never calls sys.exit.

Can an automated caller use it without guessing, and recover from failure?

  • next: L1 Errors are specific exception types, and library code never calls sys.exit. (library code exits the process: .github/actions/people/people.py)

library ladder · level 0

  • ✗ L1 Errors are specific exception types, and library code never calls sys.exit. library code exits the process: .github/actions/people/people.py
  • ✓ L2 Inputs are validated with messages naming the bad argument. pydantic/_internal/_validators.py 43 matchespydantic/_internal/_generate_schema.py 25 matches
  • ✗ L3 Timeouts and retries are configurable. no retry parameter
  • ✗ L4 Agent docs, and examples that run as tests (doctests). no agent docs
Embeddability 4 exemplary Optional features behind extras/features.

Can another system run and compose it without a human?

library ladder · level 4

Extensibility 0 absent Not yet: Internal seams: a registry, registration function or plugin manager.

Can behaviour be added without modifying the code?

  • next: L1 Internal seams: a registry, registration function or plugin manager. (no registry or registration function)

all ladder · level 0

  • ✗ L1 Internal seams: a registry, registration function or plugin manager. no registry or registration function
  • ✗ L2 The docs show how to add behaviour through those seams (even in-tree). no registration functions to document
  • ✗ L3 Out-of-tree extensions are discovered (entry points, a plugin directory, configuration) without editing the package. extensions must be added inside the package
  • ✗ L4 The extension API is versioned, has lifecycle hooks, and has example extensions. the extension API isn't versioned
Observability 0 absent Not yet: Logs through a standard logger namespace and never prints from library code.

Can an operator, or an orchestrating agent, tell whether it works and why not?

  • next: L1 Logs through a standard logger namespace and never prints from library code. (no standard logger)

library ladder · level 0

  • ✗ L1 Logs through a standard logger namespace and never prints from library code. no standard logger
  • ✓ L2 Log levels are used meaningfully (at least two levels). levels: error, info
  • ✗ L3 Optional metrics or tracing hooks. no metrics or tracing hooks
  • ✗ L4 OpenTelemetry instrumentation available. no OpenTelemetry instrumentation
Change legibility 2 partial CI runs the tests; applications pin dependencies (libraries and zero-dependency projects are exempt).

Can an agent change it safely with limited context?

  • next: L3 At least 90% of modules fit 32k tokens and no file exceeds 12k; test/source token ratio at least 0.3; at least 70% of the code is typed. (59% of modules fit 32k tokens; 10 files over 12k)

all ladder · level 2

  • ✓ L1 A README and some tests. README.mdpydantic-core/tests/benchmarks/complete_schema.py 339 test files
  • ✓ L2 CI runs the tests; applications pin dependencies (libraries and zero-dependency projects are exempt). .github/workflows/ci.yml.github/workflows/codspeed.yml
  • ✗ L3 At least 90% of modules fit 32k tokens and no file exceeds 12k; test/source token ratio at least 0.3; at least 70% of the code is typed. 59% of modules fit 32k tokens; 10 files over 12k
  • ✗ L4 No import cycles, and agent docs (AGENTS.md, CLAUDE.md, llms.txt). 3 import cycles (largest: pydantic/_internal/_config.py, pydantic/_internal/_dataclasses.py, pydantic/_internal/_decorators.py)

Make or buy

Make with agents

5,735,757 output + 143,393,925 input tokens on anthropic:claude-opus-5-5 ≈ $225.

63.7 agent-hours · 38.2 human days (38.2 steering + 0.0 rediscovering history).

$30,048 over 3 years

Buy (adopt as is)

1 min to running (via library) + 2 min to first use (via library).

30 min/year of upgrades; price $0/year.

$104 over 3 years

Buying wins below $9,981 per year. How this is estimated.

Classic COCOMO vs neosloc value

Classic COCOMO (sloccount)

79.8 KSLOC → 238.4 person-months over 20.0 months.

$2.7M

The effort to write it by hand, as COCOMO priced it in 1981.

neosloc value (neoCOCOMO)

Reproduce with agents: 23.8 PM (×0.1 of classic; 100% of behaviour captured by tests, contract and docs).

Rediscover uncaptured history: 0.0 PM (1219 fixes in 5793 commits by 835 authors).

× leverage 0.95 (integrability) × obsolescence 0.95.

$243k (21.6 PM)

What survives

9% of the classic figure

0% of the replacement cost lives only in code and history.

How the value is estimated.

History

2026-10-07: 1.622026-10-07: 1.622026-10-07: 1.622026-10-07: 1.622026-10-12: 1.62
EvaluatedCommitneoslocIndexMake or buy
2026-10-123eca5695dabc0.8.01.62buy
2026-10-07895c0c285c5c0.8.01.62buy
2026-10-07895c0c285c5c0.7.41.62buy
2026-10-07895c0c285c5c0.7.31.62buy
2026-10-07895c0c285c5c0.7.21.62buy