spec-first
This is how Agent Smith is built every day, across more than a thousand specs.
A fresh reviewer checks the spec against the code before the build. A gate runs your checks before the commit.
One branch and worktree per spec. Ids come from the clock, so two agents never take the same one.
A reviewed spec has already answered the open questions, so the agent builds to its done criteria instead of coming back to ask.
One spec, one branch, one commit. A pull request carries one spec or a whole series, and every change has the same shape.
State lives in context.yaml. Restart the machine or the agent and the work picks up where it stopped.
Each choice is logged with what it beat and why, so the reasoning stays next to the code.
Cut a bigger change into a series of specs. /spec-first:deliver-series reviews all of them, builds the groups that share no file at the same time, verifies the merged tree and opens one pull request.
deliver-series 2026-10-27-8a3f5 specs · 3 groupsbuilding
atickets-provider-contractrequires nothing1
btickets-github-providerrequires a1
ctickets-jira-providerrequires a1
dsettings-provider-pagerequires nothing2
eci-restore-cacherequires nothing3
An example run of /spec-first:deliver-series. Specs that share no file are built at the same time; specs that depend on each other share a group.
Four skills take a piece of work through its own branch and worktree. Each one leaves a file behind for the next step and for whoever reads the history later.
Goal, steps, facts with file and line, what done means. No code yet.
specs/planned/{id}.yaml
Every citation resolved, then a reviewer who never saw the chat attacks it.
BUILD · CUTS · REFUSE
Contracts, code, wiring, tests, following the project's own rules.
decisions/{id}.yaml
One commit through the gate, pushed, pull request opened.
feat: … ({id})
The recordgrows with every spec, read before every step
specs/
Every spec, from planned to done, with its facts and done criteria.
decisions/
What each spec chose, what it chose over, and why.
memory/
What you told the agent once, so you don't say it twice.
/spec-first:bootstrap-project writes the contexts once from your code: stack, rules and the checks a commit must pass. After that, each spec adds one line of state.
A REFUSE costs you one spec. Without the review, the same mistake would have cost you the branch.
Spec2026-10-27-8a3f
Applies toserver
Branchspec/8a3f
Goal
Ticket providers so pipelines can pull work from GitHub and Jira.
Facts, read in the code
Pipelines take work only from the CLI todayProgram.cs:40-58
Config already carries a tracker sectionAppConfig.cs:12-19
Decision
Adapter over repository: read-mostly data, no unit of work needed.
Done when
REVIEWED · BUILDfresh reviewer · 27 OCT
review-spec · 2026-10-27-8a3f2 cuts
✓ src/Host/Program.cs:40-58 resolves
✓ src/Config/AppConfig.cs:12-19 resolves
evidence resolves: 1 spec
BUILD WITH CUTSthe spec holds once two steps change
CUT 1
Step jira needs a second auth mode the config can't hold yet. AppConfig.cs:15 Move it to its own spec.
CUT 2
Paging is missing: GitHub returns 30 issues per page. GitHubClient.cs:88 Add a done criterion.
A script opens each file and line the spec cites. A claim nobody checked in the code has to be listed as an assumption.
A new agent with no conversation history reads the code as it is, not as the spec says it is.
BUILD, BUILD WITH CUTS or REFUSE, and every finding names a file and a line.
Nobody has to write documentation afterwards. While it builds, the agent logs each choice that wasn't obvious: what it chose, what it chose over, and why. One file per spec, next to the code.
decisions/2026-10-27-8a3f.yaml
spec: 2026-10-27-8a3f decisions: - category: TradeOff chose: "Raw HttpClient for the GitHub API" over: "Octokit" reason: "Three megabytes of dependencies for four calls." - category: Architecture chose: "One ITicketProvider per tracker" over: "A generic REST mapper" reason: "Each tracker pages and authenticates differently; the mapper would leak it."
cat decisions/*.yaml is the project's reasoning, spec by spec, in the same history as the code it explains.
How the operator wants things done, a constraint outside the repo, a pointer elsewhere. One fact per file, one line in an index.
Restart the machine, restart the agent, open context.yaml. It knows what is done, what is active and what comes next.
The plugin's commit gate runs every verify stage your contexts declare when a commit names its spec. All green, it commits. The first red stage blocks it, and its output goes back to the agent.
the commit
git commit -m "feat: ticket
providers (2026-10-27-8a3f)"
The id in parentheses triggers it. -F <file>, -C and --amend --no-edit are gated too.
verify: stages
✓builddotnet build
✓testsdotnet test
✕webcd web && npm test
PASSED
every stage green
BLOCKED
the first red stage stops it, and its output goes back to the agent
# spec-first-gate.log · one line per spec commit
2026-10-27T09:14:02Z blocked 2026-10-27-8a3f /work/repo a1b2c3d web
2026-10-27T09:31:47Z passed 2026-10-27-8a3f /work/repo a1b2c3d build,tests,web
A spec commit with no line in the ledger never met the gate. A project with its own gate says gate: project, and the plugin steps aside.
Built this way
The open source agentic software factory: a ticket goes in, a verified pull request comes out. Agent Smith is entirely built with this method, and the method grew out of building it. Everything in the plugin ran there first.
1,300+specs done
1,190+decision files: chose, over, why
650+specs through the gate
Counted in the Agent Smith repository, October 2026.
In Claude Code the plugin does the bookkeeping and runs the gate. Other agents follow the same files through AGENTS.md, and a git hook runs the same gate.
/plugin marketplace add holgerleichsenring/specification-first-agentic-development
/plugin install spec-first@specification-first-agentic-development
Then run /spec-first:bootstrap-project. It reads your code and writes the contexts and rules from what it finds.
| create-spec | Writes the spec on its own branch and worktree |
| review-spec | Evidence check, then a fresh reviewer's verdict |
| apply-spec | Builds the spec up to its done criteria |
| deliver-spec | Commits through the gate and opens the pull request |
| deliver-series | Takes a whole series of specs to one pull request |
| log-decision · update-project · bootstrap-project | Decisions, upgrades, setup |
Specification-driven agentic development, in short answers.
A way of working with AI coding agents in which every change starts as a written spec in your repository. A fresh reviewer checks the spec against your code before anything is built, the agent builds it on its own branch and worktree, and the commit lands only when your checks pass.
The spec drives the work from start to finish. It names the goal, the facts it rests on with file and line, the steps and the done criteria, and the agent builds to exactly that. Spec-first agentic development is specification-driven agentic development in which the spec is reviewed against the code before the first line is written.
No. The Claude Code plugin runs the steps as commands. Codex, Cursor, Copilot or a terminal follow the same files through AGENTS.md, and a git commit-msg hook runs the same commit gate in every clone where it is installed.
You still describe the feature in your own words and talk it through. After that, the agent works from a reviewed spec instead of from the last prompt, logs each decision with its reason, and a commit lands only when your build and tests pass.
A one-sentence goal, the facts it rests on with file and line, the steps, what an upgrade costs an operator, and the done criteria. The decisions made while building go into a separate YAML file next to it.
Nothing. The method, the templates and the Claude Code plugin are MIT licensed.
Yes, and these specs aren't that precise on purpose. A spec pins down intent and architecture: what to build, why, which code it rests on and what done means. The tests and the commit gate supply the precision.
A done spec is never updated and isn't read by default. It's a change record, like a commit with its reasoning attached. Agents read the code to learn what the system does, and search the spec index when they need to know why.
The values are still plain English. The schema keeps them short and lets a script check every cited file and line. When Agent Smith moved 93 specs from Markdown to YAML, they shrank to a fifth of their size.
A plan lives in one session. A spec stays in the repository, is reviewed against the code before it's built, and its definition of done is checked when the change is committed.
All of them write a spec or a plan before the code. spec-first checks the spec against the code before the build and blocks the commit while a check is red. The comparison page at spec-first.dev/compare/ goes through it row by row.