RFD 001: The JP RFD Process

• Status: Implemented
• Category: Process
• Authors: Jean Mertz git@jeanmertz.com
• Date: 2025-07-17
• Extended by: RFD 100

Summary

This document establishes the Request for Discussion (RFD) process for the JP project. RFDs are short design documents that describe a significant change — a new feature, an architectural shift, a process change — before implementation begins. The goal is to think clearly, communicate intent, and invite feedback early.

Motivation

JP already has a docs/architecture directory with documents that describe system designs in detail. Several of these (query stream pipeline, structured output, stateful tool inquiries, wasm tools) are effectively design proposals: they describe what we intend to build, why, and how. But they lack a formal lifecycle — there is no way to distinguish a proposal under discussion from an accepted design, or to track when a document was superseded.

We want a lightweight process that:

1. Gives design documents a clear lifecycle (draft → discussion → accepted → implemented; or abandoned / superseded).
2. Lowers the barrier to proposing ideas — rough thoughts are welcome.
3. Creates a searchable record of decisions and their rationale.
4. Works naturally with our existing Git + pull request workflow.

We do not want a process that adds bureaucracy, requires approvals from committees, or discourages people from writing things down.

Principles

The RFD process is guided by a few core beliefs, drawn from the IETF's original RFC spirit and refined for a small, fast-moving open-source project.

Timely over polished

A rough document written now is more valuable than a perfect document written never. RFDs are encouraged to be concise and direct. An RFD can be a single page. Grammar and formatting matter less than clarity of thought.

"Notes are encouraged to be timely rather than polished. Philosophical positions without examples or other specifics, specific suggestions or implementation techniques without introductory or background explication, and explicit questions without any attempted answers are all acceptable. The minimum length for a note is one sentence."

— RFC 3, Steve Crocker, 1969

Opinionated with options

An RFD should propose a specific solution, not present an open-ended menu of choices. The author's job is to navigate the problem space, evaluate alternatives, and land on a recommendation. Readers should understand what is proposed, why it was chosen, and what else was considered.

Ambiguity creates unproductive discussion. If you're unsure about the solution, that's fine — state what you know, what you don't, and what you recommend given current information. Spike with code if you need to build confidence.

Small scope

Keep RFDs focused. One document, one topic. If a change has multiple independent parts, write multiple RFDs. A focused document is easier to review, easier to discuss, and leads to faster consensus.

Use the "Non-Goals" or "Future Work" sections to acknowledge related concerns you're deliberately deferring. This signals awareness without bloating the current proposal.

Permanent record

RFDs are never deleted. If an idea is abandoned, the document is marked as such with a brief explanation. If a design is superseded, the old document links to the new one. This preserves the reasoning behind past decisions and helps future contributors understand why things are the way they are.

When to Write an RFD

Write an RFD when:

• Adding a new feature that affects the architecture or public interface
• Making a significant change to the data model or event system
• Introducing a new dependency, protocol, or integration pattern
• Changing the build, release, or contribution process
• Removing a feature or deprecating an interface
• Proposing a large refactoring effort
• Any change where you want structured feedback before investing in code

Do NOT write an RFD for:

• Bug fixes
• Performance improvements with no architectural change
• Code reorganization that doesn't change behavior
• Small feature additions that fit within established patterns
• Documentation updates

When in doubt, start writing. If it turns out to be unnecessary, you'll know quickly. The cost of an unnecessary RFD is low; the cost of a misaligned implementation is high.

RFD Lifecycle

An RFD moves through the following states:

Draft → Discussion → Accepted  → Implemented ┐
                   ↘ Abandoned ↘ Superseded ◄┘

Most RFDs follow the happy path: Draft → Discussion → Accepted → Implemented. The remaining states handle the less common cases:

• Abandoned: The idea was rejected or withdrawn during discussion.
• Superseded: An accepted or implemented design was later replaced by a new RFD.

Draft

The author is actively writing the document. It may be incomplete, have open questions, or change significantly. Drafts live on a branch and are not yet ready for formal review, but early feedback from collaborators is encouraged.

Drafts do not have a permanent number. They live in docs/rfd/drafts/ and are named DNN-slug.md, where D stands for "Draft" and NN is a two-digit draft slot (D01–D99). This prevents speculative cross-draft dependencies and avoids number gaps from abandoned drafts. The permanent number is assigned when the RFD advances to Discussion.

Drafts are not published to the documentation site. The drafts/ directory is excluded from the build, so drafts exist in the repository for committing, reviewing, and iterating without the pressure of a public page. Abandoned drafts cost nothing beyond the disk space they occupy. Published RFDs must not link to drafts — any such link would fail the docs build.

Drafts cannot be superseded. If a draft is replaced by another in-progress draft or by a new RFD before it advances to Discussion, the original draft is deleted, not superseded. Supersedes-relationships only apply once an RFD has been promoted to Accepted or beyond — there is no design to preserve before that point. Drafts that the author has decided not to pursue can be either deleted (no rationale worth preserving) or abandoned via just rfd-abandon (rationale worth preserving as a record of the rejected idea); the choice is the author's.

Discussion

The RFD is complete enough to review. When the author runs just rfd-promote D01 (using the draft ID), the tooling:

1. Assigns the next available sequential permanent number.
2. Moves the file from docs/rfd/drafts/D01-slug.md to docs/rfd/042-slug.md (for example).
3. Updates the heading in the file.

A pull request is opened to merge the document into main. Discussion happens on the pull request. The author incorporates feedback and iterates on the document.

There is no fixed timeline for discussion. For most RFDs, a few days should suffice. If no one has reviewed your RFD after 48 hours, ask someone directly. If discussion stalls, a synchronous conversation (call, chat) can help break the deadlock.

Accepted

Discussion has converged and the pull request is merged. The RFD represents the agreed-upon direction. Implementation can begin.

An accepted RFD is not immutable. If implementation reveals issues, update the document. For minor corrections, edit in place. For significant changes, write a new RFD that supersedes the original.

Implemented

The feature or change described in the RFD has been fully implemented. This is a bookkeeping state — it signals that the document describes the current system, not just a plan.

Superseded

The design in this RFD has been replaced by a newer RFD. The original document remains in the repository as a historical record. Its metadata is updated with a Superseded by link pointing to the replacement, and the new RFD carries a Supersedes link pointing back.

Superseded is distinct from Abandoned: a superseded RFD was accepted and may have been partially or fully implemented, but a later design replaced it. An abandoned RFD was never accepted or implemented.

An RFD can be superseded from either the Accepted or Implemented state. Drafts cannot be superseded — a draft replaced before promotion is deleted (see Draft above).

Abandoned

The idea was considered and deliberately set aside. The document remains in the repository with a brief note explaining why. Common reasons: the problem was solved differently, priorities changed, or the approach turned out to be infeasible.

An abandoned RFD opens with a standard notice block:

> [!IMPORTANT]
> This RFD is **Abandoned**.
>
> {one-line reason}
>
> {what was carried forward and where it lives now, if anything.}
>
> The original text below is preserved for historical context.

If the RFD was split into other RFDs, name the active successors. If portions were superseded by a later RFD, name it. If nothing was carried forward, omit the second paragraph.

Document Format

Filename

Drafts live under docs/rfd/drafts/ and use a DNN prefix (D for Draft, NN a two-digit slot):

docs/rfd/drafts/D01-short-title.md

When promoted to Discussion, the tooling assigns a permanent number and moves the file up to docs/rfd/:

docs/rfd/042-short-title.md

• DNN is used for drafts; a zero-padded sequential number (001, 002, ...) is assigned at Discussion.
• short-title is a lowercase, hyphen-separated slug. Keep it short but descriptive.
• Numbers are never reused. If an RFD is abandoned, its number is retired.

Templates

Every RFD has a Category that describes its purpose. Each category has a corresponding template:

Category	Template	Use when
Design	000-design-template.md	Proposing a feature or
		architectural change that needs a
		design.
Decision	000-decision-template.md	Recording a decision — a technology
		choice, a convention, a policy.
Guide	000-guide-template.md	How-tos, reference material, and
		contributor-facing documentation.
Process	000-process-template.md	How the project operates:
		workflows, policies, values.

All categories share the same numbering scheme, directory, lifecycle, and review process. The difference is purpose: a Design has a full design section and implementation plan; a Decision has concise context and consequences; a Guide or Process document is free-form.

The templates are starting points, not constraints. Structure the document however it reads best. The only requirement is the metadata header (Status, Category, Authors, Date) so the tooling and lifecycle work. Delete the template sections that don't apply, add sections that do, or write something entirely free-form.

To create a new draft:

just rfd-draft design "My Feature Title"        # design proposal
just rfd-draft decision "Use TOML for Config"   # decision record
just rfd-draft guide "Attachment Handler Guide" # how-to / reference
just rfd-draft process "Release Process"        # workflow / policy

This copies the appropriate template to docs/rfd/drafts/D01-my-feature-title.md (or the next available draft slot), fills in the title, author, date, and category, and sets the status to Draft. The draft prefix is replaced with a permanent number and the file moves up to docs/rfd/ when the RFD is promoted to Discussion.

Document sections

Not all sections are required for every RFD — omit sections that genuinely don't apply — but think twice before skipping one. Every section can be brief. A one-sentence Alternatives section is better than no Alternatives section.

Metadata header

All categories use the same metadata:

- **Status**: Draft | Discussion | Accepted | Implemented | Superseded | Abandoned
- **Category**: Design | Decision | Guide | Process
- **Authors**: Name <email> (or GitHub handle)
- **Date**: YYYY-MM-DD
- **Extends**: RFD NNN (if this RFD builds on another)
- **Extended by**: RFD NNN (if another RFD builds on this one)
- **Requires**: RFD NNN (if this RFD depends on another)
- **Required by**: RFD NNN (if another RFD depends on this one)
- **Supersedes**: RFD NNN (if applicable)
- **Superseded by**: RFD NNN (if applicable)

Implementation progress is tracked in tickets, not in the metadata header. When just rfd-promote advances an RFD from Discussion to Accepted, it offers to turn each phase of the Implementation Plan into a ticket carrying Implements: NNN. The phases are read out of the document by jp, so the tickets match the plan rather than a fixed template, and whoever accepts the prompt reviews them before they land.

Accepting an RFD records an agreed direction, not a commitment to start building, which is why the tickets are offered rather than created. Decline the prompt and file them later with just ticket-create when the work actually starts.

A handful of older RFDs still carry a Tracking Issue field pointing at a GitHub issue. The field is retired: nothing writes it, and the lifecycle recipes only read it to remind you to close the issue when such an RFD is superseded or abandoned.

The Extends and Extended by fields capture design-lineage relationships: "this RFD builds on that one's design." They are maintained by just rfd-extend. Unlike Supersedes, an extending RFD builds on its predecessor — the original remains valid and in effect.

The Requires and Required by fields capture hard dependencies: "this RFD cannot be Accepted (or Implemented) until that one is." They are maintained by just rfd-require.

Requires exists only to gate promotion on an unbuilt dependency. Once a target reaches Implemented, the dependency is satisfied for good and the link serves no further purpose, so a Requires entry is never recorded against an already-Implemented RFD. When an RFD itself reaches Implemented, just rfd-promote strips its Requires entries and the matching Required by back-links on the targets; by that point the gate guarantees every target is already Implemented or Superseded. The docs build rejects any published RFD that lists a Requires on an Implemented target. Use Extends instead when the relationship is design lineage worth recording past implementation.

Both relationships participate in the same gate. An extension is a kind of dependency (Extends ⊆ Requires) — if A extends B, then A also depends on B. So rfd-promote and the docs build enforce both fields uniformly:

• Promoting an RFD from Discussion to Accepted requires every entry in Requires or Extends to be at status Accepted, Implemented, or Superseded.
• Promoting from Accepted to Implemented requires every entry to be at status Implemented or Superseded.

This lets RFDs be designed in parallel (multiple RFDs in Accepted at once) while preventing claims of design completion that rest on an unbuilt foundation. Cycles in the union graph are refused at write time and by the docs build.

Don't list the same target in both Extends and Requires. Extension implies the dependency, so listing the same target twice is redundant. Pick Extends when the relationship is design lineage; pick Requires when it is only execution prerequisite. The recipes refuse to write a duplicate, and the docs build refuses to publish one.

Drafts may participate in the dependency graph, but back-links from non-draft RFDs are suppressed: a published RFD never lists a draft under Required by or Extended by. When a draft is promoted, missing back-links on its dependency targets are filled in automatically by rfd-promote.

Writing Style

• Use present tense. "This RFD describes..." not "This RFD was created to describe..."
• Be direct. Avoid hedging language like "it seems" or "probably" or "it might be worth considering." State what you propose and why.
• Use concrete examples. A code snippet or data flow diagram is worth a paragraph of abstract description.
• Define terms. If you introduce a concept, define it where it first appears.
• Keep it short. If an RFD exceeds 5-6 pages (roughly 2000 words), consider whether it can be split into smaller proposals.

Process

Creating an RFD

1. Create a branch for your work.
2. Run just rfd-draft CATEGORY Your Title to generate the file from the appropriate template. Category is one of: design, decision, guide, process. The file is created under docs/rfd/drafts/ with a DNN prefix (e.g. docs/rfd/drafts/D01-your-title.md).
3. Fill in the sections. Write your proposal.
4. Push your branch. Iterate until you're ready for feedback.

Opening for Discussion

1. Run just rfd-promote D01 to advance the status to Discussion. This assigns a permanent number and renames the file.
2. Open a pull request to merge your branch into main.
3. Tag reviewers — people with context on the problem area.
4. Engage with feedback. Update the document as the discussion evolves.

Accepting an RFD

1. When discussion converges, run just rfd-promote NNN to advance the status to Accepted. The promotion is gated on the RFD's Requires and Extends fields: every entry in either field must be at status Accepted, Implemented, or Superseded.
2. Accept the prompt to file a ticket per implementation phase, or decline it and file them when the work starts.
3. Merge the pull request.

After Acceptance

• Minor updates: Edit the document directly on main via a standard pull request. No new RFD number needed.
• Significant changes: Write a new RFD that supersedes the original.
• Implementation complete: Run just rfd-promote NNN to advance the status to Implemented.
• Design extended: When a new RFD builds on this one, run just rfd-extend NNN MMM to record the relationship in both documents.
• Design depended on: When a new RFD requires this one as a hard prerequisite, run just rfd-require NNN MMM (NNN requires MMM) to record the dependency. The relationship gates promotion of the dependent RFD.
• Design superseded: Write a new RFD, then run just rfd-supersede NNN MMM to mark the old RFD as superseded and cross-link both documents.
• Idea abandoned: Run just rfd-abandon NNN "reason" to mark the RFD as abandoned with an explanation.

Tooling

All RFD commands are in the rfd group. Run just --list --group rfd to see them.

Command	Description
just rfd-draft CATEGORY TITLE	Create a new draft under drafts/.
just rfd-promote NNN	Advance status. Draft → Discussion
	assigns number; Discussion → Accepted
	offers phase tickets.
just rfd-extend NNN MMM	Record that RFD MMM extends RFD NNN,
	updating both. Accepts draft IDs (DNN)
	on either side.
just rfd-require NNN MMM	Record that RFD NNN requires RFD MMM,
	updating both. Cycles are refused.
	Accepts draft IDs (DNN) on either side.
just rfd-supersede NNN MMM	Mark RFD NNN as superseded by RFD MMM,
	updating both.
just rfd-abandon NNN REASON	Mark RFD NNN as abandoned with the given
	reason.
just rfd-renumber NNN [MMM]	Renumber an RFD (draft or published) to
	MMM, or the next available id, and
	rewrite all cross-references.
just rfd-grep TERM	Search across all RFD documents using
	rg.
just rfd-list [CATEGORY]	List all RFDs (including DNN-prefixed
	drafts), optionally filtered by
	category.

Relationship to Architecture Documents

The existing docs/architecture/ directory contains detailed technical descriptions of implemented systems. These serve a different purpose than RFDs:

	RFDs (docs/rfd/)	Architecture Docs (docs/architecture/)
Purpose	Propose a change	Describe the current system
Lifecycle	Draft → Accepted → Implemented	Living documents, updated as the system
		evolves
Audience	Contributors deciding what to build	Contributors understanding what exists
Scope	A specific change or feature	A subsystem or cross-cutting concern

The typical flow: an RFD proposes a design, gets accepted, and is implemented. Once implemented, the relevant architecture documents are updated to reflect the new state of the system. The RFD remains as a historical record of the decision.

Over time, some existing architecture documents may be retroactively referenced by RFDs, or new architecture documents may be created as companions to accepted RFDs. The two directories complement each other.

FAQ

What if I'm not sure about the solution?

Write what you know. State the options you see and which one you lean toward. Use the "Risks and Open Questions" section to flag uncertainty. A draft with acknowledged unknowns is more useful than no document at all.

If you need to experiment first, do that. Write the RFD after you've spiked and have a clearer picture.

How detailed should the design section be?

Detailed enough that a reviewer can evaluate the approach without reading the implementation code. Not so detailed that it becomes the implementation spec. RFDs describe the "what" and "why" at an architectural level; the code is the "how" at an implementation level.

For JP specifically, the existing architecture documents provide a good reference for the level of detail expected: design goals as tables, data flow descriptions, component responsibilities, migration paths.

Can I update an accepted RFD?

Yes. For small corrections (typos, clarifications, minor adjustments discovered during implementation), edit the document directly. For changes that alter the fundamental approach, write a new RFD.

What about the existing architecture documents?

They stay where they are. The architecture directory describes the system as it is. The RFD directory captures proposals and decisions. Both are valuable. See Relationship to Architecture Documents.

Do I need approval to merge an RFD?

Follow the project's normal pull request process. An RFD should be reviewed by at least one other contributor with relevant context before merging. The goal is consensus, not a formal sign-off process.

What if my document doesn't fit either template?

Use whatever structure makes sense. The templates are suggestions to help you get started, not a format you must follow. Policy documents, values statements, process guidelines — these have their own natural shape. The only hard requirement is the metadata header (Status, Authors, Date) at the top of the file, so the lifecycle tooling works. See RFD 002 for an example of a free-form RFD.

---

Implementation Plan

This RFD is its own implementation. The steps are:

1. Create the docs/rfd/ directory.
2. Add 000-design-template.md, 000-decision-template.md, 000-guide-template.md, and 000-process-template.md.
3. Add just tasks for the RFD lifecycle: rfd-draft, rfd-promote, rfd-supersede, rfd-abandon, and rfd-grep.
4. Add this document as 001-jp-rfd-process.md.
5. Merge via pull request after discussion.
6. Future RFDs follow the process described here.