Product overview

Source: docs/product/overview.mdVerified 2026-10-15

The problem

An AI coding agent starts every session from zero. Decisions made last week, the reason one library was rejected, the bug that took an afternoon: none of it survives the session unless someone writes it down where the next session will look. At the same time, the agent's defaults are not the habits of a careful engineer: it commits without running the right checks, claims "fixed" without evidence, reviews its own code, installs dependencies unasked and forgets the user's language after compaction.

What developer-os is

A Claude Code plugin, installed from this repository as a plugin marketplace, that adds two things:

  1. Long-term memory in an Obsidian vault. Six brain-* skills let the agent search the vault before a task, capture decisions and lessons into the vault's _inbox/ while it works, and help the user review, garden and consolidate what was captured. The vault is plain Markdown; the agent reads and writes it with its own file tools. The user approves everything that becomes knowledge.
  2. An engineering operating model. Core rules loaded into every session, workflow skills (review, spec, audit, backlog, PR cycle, triage, deploy checklist, docs sync, session wrap-up), conventions skills, five specialist agents, four output styles and a small Bash guard.

The plugin is Markdown and JSON plus three Python 3 hook scripts (standard library only). There is no server, database, index, binary, background job or network call in the plugin itself.

Who it is for

Product principles

  1. The human approves what becomes knowledge. The agent creates notes only in _inbox/ with status: inbox; it edits, moves or archives any other note only after the user names it.
  2. Archive, never delete. Retired, merged and rejected notes move to _archive/ with status: archived.
  3. Propose, then act on a named yes. Every skill that changes files, git state or anything outside the machine stops at a defined point and asks; silence or "looks good" is not approval. Each action skill states its ## Stop point and ## Never sections, and scripts/check.sh fails without them.
  4. Evidence over declaration. Findings carry a file and line, reviews rate on one P0–P3 scale, blocking findings go to a fresh verifier, and "works" needs test output.
  5. One fact in one place. Code facts and plans stay in the repository; lessons that outlive a repository go to the vault; user preferences go to auto-memory or ~/.claude/CLAUDE.md (workflow.md, Where knowledge belongs).
  6. Small fixed cost per session. The core rules are capped at 8192 bytes and the skill descriptions at 12000 characters in total, because both load into every session.
  7. No runtime. Anything that would need a process, a schedule or the network is either left to Claude Code itself (routines, the marketplace) or not built.

Scope

In scope for the first release:

ComponentCountSource
Brain skills6plugin/skills/brain-*
Engineering skills13plugin/skills/
Session skills2 (session-retro, skill-usage)plugin/skills/
Conventions skills7 (*-conventions)plugin/skills/
Agents5plugin/agents/
Output styles4plugin/output-styles/
Hooks3 (SessionStart, Stop, PreToolUse on Bash)plugin/hooks/hooks.json
Core rules1 fileplugin/rules/core.md
Vault template1plugin/vault-template/
Project template1templates/project/ (in the repository, not in the installed plugin)
Eval casesone directory per caseevals/ (in the repository, not in the installed plugin)

The README lists every component by name (What's inside); the functional specification gives each one's contract.

Out of scope, on purpose:

How it is distributed

Status

Release 0.2.4; installs follow main, which moves only on a release. See CHANGELOG.md for what each release changed.

Primary risks

RiskEffectMitigation today
Rules and descriptions grow until they crowd the contextEvery session pays for themByte and character caps in scripts/check.sh
A skill's description does not match how the user asksThe skill never firesTrigger phrases in descriptions, trigger-* evals, the skill-usage report
An update changes the hooks, which run in every sessionUnreviewed code runs in the user's shellHooks are short Python 3 scripts, standard library only, with tests; SECURITY.md asks users to review each update's diff
An unreviewed agent note is treated as factWrong knowledge spreadsstatus: inbox until approved; brain-search marks inbox claims as unreviewed
Vault content reaches the model providerClient context leaves the diskRedaction in brain-capture; privacy rules in workflow.md