Metadata-Version: 2.4
Name: solodeveling
Version: 1.0.0
Summary: Single-agent-first software delivery protocol and portable skill suite
Project-URL: Homepage, https://github.com/ohmiler/solodeveling
Project-URL: Source, https://github.com/ohmiler/solodeveling
Project-URL: Issues, https://github.com/ohmiler/solodeveling/issues
Project-URL: Documentation, https://github.com/ohmiler/solodeveling#readme
Author: ohmiler
License-Expression: Apache-2.0
License-File: LICENSE
Keywords: agent-skills,coding-agents,secure-sdlc,solo-developer
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Programming Language :: Python :: 3.14
Classifier: Topic :: Software Development :: Quality Assurance
Requires-Python: >=3.10
Requires-Dist: jsonschema<5,>=4.23
Requires-Dist: pyyaml<7,>=6.0
Provides-Extra: dev
Requires-Dist: build<2,>=1.2; extra == 'dev'
Requires-Dist: cyclonedx-bom==7.3.0; extra == 'dev'
Requires-Dist: pyinstaller==6.21.0; extra == 'dev'
Requires-Dist: pytest==9.1.1; extra == 'dev'
Requires-Dist: tomli<3,>=2.2; (python_version < '3.11') and extra == 'dev'
Provides-Extra: native
Requires-Dist: pyinstaller==6.21.0; extra == 'native'
Description-Content-Type: text/markdown

# Solodeveling

[English](README.md) | [ภาษาไทย](README.th.md)

> **One agent. The right amount of process. From a tiny fix to a verified release.**

Solodeveling is a single-agent-first software delivery protocol for solo developers.
One public adaptive skill covers discovery, planning, implementation, diagnosis,
verification, security, release, and maintenance. Small, safe work stays inline.
Resumable work gains one current record, while risky boundaries activate stronger
authority, recovery, and evidence profiles independently of task size.

Use the same canonical skill with Codex, Claude Code, Cursor, and generic Agent Skills
clients. Subagents may still be used when useful, but Solodeveling does not
require a multi-agent team to complete the ordinary workflow.

**Status:** Alpha releases are published on
[npm](https://www.npmjs.com/package/solodeveling),
[PyPI](https://pypi.org/project/solodeveling/), and as an
[immutable GitHub Release](https://github.com/ohmiler/solodeveling/releases/latest).
Codex and Claude Code have each passed one bounded representative live scenario.
Cursor has structural adapter evidence only because `cursor-agent` was unavailable.
Tier 1 remains unverified until the full core scenario matrix passes.

## Quick start

Node.js users need one command:

~~~console
npx solodeveling install
~~~

For a permanent command:

~~~console
npm install -g solodeveling
solodeveling install
~~~

Python users can run it without a permanent install or install it as a tool:

~~~console
uvx solodeveling install
uv tool install solodeveling
pipx install solodeveling
~~~

All channels expose one public command: `solodeveling`. See
[Installation](docs/installation.md) for prerequisites, platform support, upgrades,
and integrity controls.

## Why Solodeveling

- **Single-agent-first:** keep one accountable primary agent instead of requiring
  delegation, role hand-offs, or parallel workers for ordinary delivery.
- **Quiet for small work:** safe inline work can use zero required questions, zero
  project-memory writes, and verification proportional to its impact.
- **Direct for read-only work:** ordinary Q&A, explanation, review, status, diagnosis,
  and inline advice inspect only what the claim needs and skip lifecycle artifacts.
- **Durable when it matters:** resumable work keeps one current record; guarded
  boundaries add explicit authority, recovery, evidence, and release controls.
- **Adaptive coverage:** intent, persistence, software effects, and assurance are
  selected independently behind one natural-language entry point.
- **Project-local and portable:** install the same skill into the runtimes detected
  in the repository without searching for or modifying global agent installations.

| Adaptive axis | Choices | What it controls |
| --- | --- | --- |
| Intent | Observe, explore, deliver, diagnose, operate | The next useful action |
| Persistence | Inline or tracked | Zero artifacts or one resumable current record |
| Effect | Content, UI, service, data, platform, AI/ML | Relevant technical checks and risks |
| Assurance | Routine or guarded | Authority, recovery, integrity, and evidence gates |

These axes are internal; users do not choose a workflow. A tiny authorization change
can be guarded, while a large documentation refactor can be tracked but routine.
References load only from observable signals. State stays a compact dashboard, WORK
is the current execution record, and separate evidence or decisions are added only
when guarded, audited, or necessary for resumption.

Inline routine work uses inspect, change, focused proof, and report with no lifecycle
artifact. Tracked work persists `active` before implementation and closes once after
reconciled verification. Roadmap updates occur only when priority, milestone,
ordering, or deferred direction changes.

### With and without Solodeveling

A coding agent can keep a `TODO.md` or write next-session notes without installing
any skill. Solodeveling does not make the model inherently smarter; it packages a
repeatable delivery policy so the user does not have to reconstruct it in every
prompt or session.

| Concern | Without a shared skill | With Solodeveling |
| --- | --- | --- |
| Coordination | Instructions and working conventions depend on the current prompt, agent, and session | One project-local adaptive skill provides the same contract on each supported runtime |
| Continuity | Notes can use any format the user or agent chooses | Compact state records only resumable context, with one owner for each durable fact |
| Small work | Can have almost no ceremony when the agent makes the right call | Inline routine work explicitly requires no lifecycle artifacts |
| Completion | The prompt or session determines what evidence is enough | Completion claims require recent verification scoped to the changed boundary |
| Risky work | Security, authority, recovery, and release checks must be supplied by the prompt or another process | Guarded assurance activates explicit security, recovery, provenance, and authorization gates |

Not every project needs Solodeveling. A small, low-risk, single-session project may
be better served by a clear prompt and a `TODO.md`. Solodeveling earns its overhead
when repeatable decisions, cross-session continuity, or risk-sensitive verification
matter. The testable claim is not that a skill makes the model smarter, but that a
shared protocol reduces repeated coordination and missed gates without burdening
small work.

## How it compares

These projects have different defaults rather than a universal winner. The table
summarizes their public documentation as of July 2026; it is not a controlled speed
or quality benchmark.

| Project | Documented default | Strongest fit | Main contrast with Solodeveling |
| --- | --- | --- | --- |
| **Solodeveling** | One primary agent with risk-scaled workflow and memory | Solo maintainers balancing tiny fixes with security, release, and maintenance work | Quick work may stay ephemeral; compact state expands into auditable evidence only when needed |
| [Superpowers](https://github.com/obra/superpowers) | Complete development methodology built from composable skills | Developers who want modular engineering skills, planning, review, and subagent-oriented paths where supported | Solodeveling keeps one accountable primary agent as the ordinary default and does not require delegation |
| [GSD](https://github.com/gsd-build/get-shit-done) | Lightweight meta-prompting, context engineering, and spec-driven execution | Longer builds where countering context degradation is the main concern | Solodeveling organizes persistence and verification around task risk and the full delivery lifecycle |
| [GitHub Spec Kit](https://github.github.com/spec-kit/) | Specification-centered flow: Spec, Plan, Tasks, Implement | Work where the specification should remain the primary source of truth across many agent integrations | Solodeveling allows safe Quick work to skip a persistent spec while retaining escalation paths |
| [BMAD Method](https://github.com/bmad-code-org/BMAD-METHOD) | Scale-adaptive agile workflows with 12+ specialist roles and 34+ workflows | Users who want guided agile personas, modules, and a broad workflow ecosystem | Solodeveling favors one primary agent and a smaller common vocabulary across the lifecycle |

### What faster means here

Solodeveling optimizes **workflow overhead**, not unmeasured model intelligence. It
does not yet claim to complete coding tasks faster than the alternatives above. A
fair claim requires the same repository, model, prompt, task set, and repeated runs.
The repository now includes a
[preregistered 18-run Solodeveling/Superpowers pilot](docs/comparative-benchmark.md)
with exact source and runtime pins, hidden correctness checks, offline planning, and
a separate live-authorization gate. Pilot 1 failed before inference; Pilot 2 did
not activate either methodology; and Pilot 3 could not mutate through the broken
Windows sandbox. These are invalid measurement runs, not a comparative result.
A separate [feedback measurement plan](docs/measurement.md) preregisters a
correctness-gated 0.1.1-versus-0.1.2 pilot and a local twenty-task field scorecard.
The same plan records the completed, correctness-first 30-call comparison between
pinned Solodeveling 0.3.1 and the one-skill 1.0.0 candidate. Both arms passed all
15 tasks, including guarded readiness, while elapsed time, token, and tool-call
differences stayed below the preregistered materiality thresholds. This supports
preserved correctness in that small corpus, not a speed or universal-quality
advantage. A twenty-task adaptive 1.0 internal dogfood cohort is collecting
observational evidence separately; even a passing field gate permits review only.

| Evidence from this repository | Observed result | What it supports |
| --- | --- | --- |
| Ephemeral documentation dogfood | Zero user questions and no Quick-task memory artifacts | Small clear work can avoid coordination and persistence ceremony |
| Memory-only GitHub CI | Focused validation jobs completed in 18-20 seconds | Memory updates do not trigger package, native, or full test jobs |
| Docs-only GitHub CI | Complete Python regression job completed in 23-33 seconds | Documentation retains meaningful verification without the cross-platform package matrix |
| Bounded live portability scenario | Codex and Claude Code each scored 1.0 on one Quick scenario | The canonical behavior can transfer across two available runtimes; it does not establish full Tier 1 support |
| Pinned 0.3.1-versus-1.0.0 pilot | Both versions passed 15/15; overhead deltas were below materiality thresholds | The 1.0.0 candidate preserved correctness in this corpus; no speed advantage is established |

See [EVIDENCE-006](.solodeveling/evidence/EVIDENCE-006.md),
[EVIDENCE-019](.solodeveling/evidence/EVIDENCE-019.md), and
[EVIDENCE-025](.solodeveling/evidence/EVIDENCE-025.md) for historical methods and
limitations. The completed adaptive pilot is recorded in
[EVIDENCE-058](.solodeveling/evidence/EVIDENCE-058.md).

## Who it is for

Choose Solodeveling when you:

- maintain a project alone or with one primary coding agent;
- want tiny fixes to stay lightweight without losing discipline on risky work;
- need continuity across sessions without accumulating a document for every action;
- own both implementation and release or maintenance responsibilities; or
- switch between Codex, Claude Code, Cursor, or another Agent Skills client.

Another approach may fit better when you want specialized personas and multi-agent
coordination as the default, require a formal specification artifact for every
change, or need a mature ecosystem with a much larger behavioral evidence base.

## Automatic project installation

The ordinary workflow has no required options:

~~~console
solodeveling install
solodeveling check
solodeveling uninstall
~~~

Solodeveling first reuses any installation it already manages. Otherwise it detects
project-local Codex/Agent Skills, Claude Code, and Cursor directories. It installs
every distinct runtime found, and defaults to the standard `.agents/skills` path when
the project has no runtime marker. It never searches global executables or writes
outside the current project.

Installation validates the adaptive skill and activation manifest, preflights every detected target, rejects
symlinks and path traversal, refuses unmanaged collisions, copies files atomically,
and records managed hashes. Check detects missing, changed, or unexpected managed
files. Uninstall removes only unchanged managed files and has no force-delete mode.

Start a fresh agent session after first installation, then invoke `$solodeveling` in
Codex, `/solodeveling` in Claude Code or Cursor, or the client-defined invocation in
another Agent Skills runtime.

### Advanced overrides

Automation and unusual workspaces may still select an exact runtime or another
project. `--dry-run` remains available as an optional preview, not a required step:

~~~console
solodeveling install --runtime claude-code --project-root PATH
solodeveling install --runtime cursor --dry-run
~~~

## Other commands

The install flow stays short; project memory, tracked-work helpers, and evaluation
remain explicit:

~~~console
solodeveling init
solodeveling validate .
solodeveling work evidence . WORK-033 --claim "Focused tests pass" --method "Automated test" --result passed --scope "Lifecycle helper"
solodeveling work transition . WORK-033 verifying
solodeveling work archive . WORK-033 --next-action "Select the next priority"
solodeveling eval probe
solodeveling version
~~~

The work helper validates memory before and after each operation, rejects invalid
transitions, updates work and state together, and rolls back a partial write. Standard
work reuses its evidence file automatically; audited work with multiple evidence
files can select one with `--evidence-id`.

Live evaluation can consume model-service usage or API credits, so review
[Cross-agent evaluation](docs/cross-agent-evaluation.md) before authorizing a live
run.

## Security and support boundaries

The npm launcher has no runtime dependencies and no install lifecycle script. On an
explicit invocation it selects an exact-version native artifact, downloads only from
the versioned GitHub Release, verifies its bundled size and SHA-256, caches it, and
executes it without a shell. Windows, macOS, and Linux on x64 and arm64 are release
targets only after their CI builds and native smoke tests pass. Unsupported targets
fail closed and show the Python-tool fallback.

Solodeveling applies risk-scaled verification and Secure SDLC guidance, but installing
it does not make a project categorically secure or compliant. Hashes detect byte
changes; they do not by themselves prove publisher identity. No telemetry is
collected and no publishing credentials are included.

No tag, GitHub Release, npm package, or PyPI package is created by ordinary CI.
External publication requires a reviewed source revision and separate explicit
authorization.

## License

Apache-2.0. See [LICENSE](LICENSE).
