RuleStack

Configs

Stacks

Compare

Diff

RuleStack

Configs

Stacks

Compare

Diff

Read API

RuleStack

Configs

Stacks

Compare

Diff

Read API

Configs/agentjido/jido/diff

Two files, one repository

agentjido/jido ships 1 format across 3 indexed files. The question worth asking is whether the second one says anything the first does not.

A · AGENTS.md · 361 wordsB · .spec/AGENTS.md · 219 words
What each file covers, counted
DimensionSharedOnly in AOnly in BOverlap
Sections01030%
Commands0660%
Section tags41167%

What each file covers

Sections

0 shared · 10 only in A · 3 only in B
  • − AGENTS.md - Jido Guide
  • − Intent
  • − Runtime Baseline
  • − Commands
  • − Work Management
  • − Architecture Snapshot
  • − Standards
  • − Testing and QA
  • − Release Hygiene
  • − References
  • + `.spec` Agent Guide
  • + First Read
  • + Working Rules

Commands

0 shared · 6 only in A · 6 only in B
  • − mix test
  • − mix test --include flaky
  • − mix test --cover
  • − mix q
  • − mix quality
  • − mix docs
  • + mix spec.prime --base HEAD
  • + mix spec.next
  • + mix spec.next --bugfix
  • + mix spec.check --base ...
  • + mix spec.validate --debug
  • + mix spec.status

Section tags

4 shared · 1 only in A · 1 only in B
  • − deployment
  • + agent-behaviour
  •   test
  •   code-style
  •   git-pr
  •   do-not

Line diff

+24 added−54 removed7 unchanged11.5% identical
agentjido/jido · AGENTS.md
@@ −1 @@
1# AGENTS.md - Jido Guide
2 
3## Intent
4Build reliable agent systems by separating pure decision logic from runtime side-effect execution.
5 
6## Runtime Baseline
7- Elixir `~> 1.18`
8- OTP `27+` (release QA baseline)
9 
10## Commands
11- `mix test` (default alias excludes `:flaky`)
12- `mix test --include flaky` (full suite)
13- `mix test --cover` (coverage gate)
14- `mix q` or `mix quality` (`format --check-formatted`, `compile --warnings-as-errors`, `credo`, `dialyzer`)
15- `mix docs` (local docs)
16 
17## Work Management
18This project tracks work with `bw` (Beadwork), which persists plans, progress,
19and decisions to git so they survive compaction, session boundaries, and context
20loss.
21 
22- Always run `bw prime` before starting implementation work.
23- Use one Beadwork issue per branch/PR unless the issue is explicitly split.
24- Start work with `bw start <id>` and record validation, skip decisions, PR
25 links, and CI outcomes with `bw comment <id> "..."`
26- Completing a task includes committing, pushing, closing the Beadwork issue,
27 and running `bw sync`.
28- For architecture-review findings, prefer the portable helper:
29 `scripts/bw-codex-loop.sh`.
30 
31## Architecture Snapshot
32- `Jido.Agent`: pure agent module with immutable state and `cmd/2`
33- `Jido.AgentServer`: GenServer runtime for directives, lifecycle, and message flow
34- `Jido.Agent.Directive.*`: typed effect descriptors (`Emit`, `SpawnAgent`, `StopChild`, etc.)
35- `Jido.Agent.StateOp.*`: internal state transition operations applied by strategy layer
36- Plugins/sensors provide capability composition without coupling core agent logic
37 
38## Standards
39- Keep `cmd/2` pure: same input => same `{agent, directives}` output
40- Keep directives for external effects only; do not rely on directives for state mutation
41- Use **Zoi-first** schemas for new agent/plugin/signal contracts
42- Preserve tagged tuple and structured error contracts at public boundaries
43- Keep cross-agent communication on signals/directives, not ad-hoc process messages
44 
45## Testing and QA
46- Prefer pure agent tests first, then AgentServer/runtime integration tests
47- Use helpers from `test/AGENTS.md` (`JidoTest.Case`, `JidoTest.Eventually`) for async assertions
48- Avoid `Process.sleep/1` in tests; assert eventual state/event behavior
49 
50## Release Hygiene
51- Keep semver ranges stable (`~> 2.0` for Jido ecosystem peers)
52- Use Conventional Commits
53- Do not modify `CHANGELOG.md`; release notes are generated from Git history during release, so keep changes focused on proper Conventional Commits.
54 
55## References
56- `README.md`
57- `usage-rules.md`
58- `guides/`
59- `test/AGENTS.md`
60- https://hexdocs.pm/jido
61 
agentjido/jido · .spec/AGENTS.md
@@ +1 @@
1# `.spec` Agent Guide
2 
3Use this folder to maintain authored Spec Led Development subjects and generated state.
 
4 
5<!-- covers: spec.workspace.agents_present spec.workspace.agent_prime_context -->
 
 
6 
7## First Read
 
 
 
 
 
8 
91. Read `.spec/README.md`.
102. Read `.spec/decisions/README.md` and any ADRs that affect the subject you are changing.
113. Read the current `.spec/specs/*.spec.md` files before editing.
 
12 
13## Working Rules
 
 
 
 
 
 
 
14 
15- Keep one subject per file.
16- Put normative statements in `spec-requirements`.
17- Add `spec-scenarios` only when `given` / `when` / `then` improves clarity.
18- Add `spec-meta.decisions` only when a subject depends on a durable cross-cutting ADR.
19- Keep ADRs in `.spec/decisions/*.md` for cross-cutting policy only.
20- Prefer targeted command verifications for behavioral proof.
21- Use file-backed verifications only when the target can carry stable `covers:` markers for every covered id.
22- Keep verification targets repository-root-relative.
23- Use Git history and pull requests as the change log; keep `.spec` current-state only.
24- At session start, run `mix spec.prime --base HEAD`.
25- After code, docs, or tests change, run `mix spec.next`.
26- For bug fixes, prefer `mix spec.next --bugfix`.
27- If next says `needs subject updates`, update the named subject before you finish.
28- If next says `ready for check`, move to `mix spec.check --base ...`.
29- Use `mix spec.validate --debug` when you need low-level verifier output.
30- Run `mix spec.status` when you need coverage or weak-spot summaries.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
31 
@@ −1 +1 @@
1−# AGENTS.md - Jido Guide
1+# `.spec` Agent Guide
22  
3−## Intent
4−Build reliable agent systems by separating pure decision logic from runtime side-effect execution.
3+Use this folder to maintain authored Spec Led Development subjects and generated state.
54  
6−## Runtime Baseline
7−- Elixir `~> 1.18`
8−- OTP `27+` (release QA baseline)
5+<!-- covers: spec.workspace.agents_present spec.workspace.agent_prime_context -->
96  
10−## Commands
11−- `mix test` (default alias excludes `:flaky`)
12−- `mix test --include flaky` (full suite)
13−- `mix test --cover` (coverage gate)
14−- `mix q` or `mix quality` (`format --check-formatted`, `compile --warnings-as-errors`, `credo`, `dialyzer`)
15−- `mix docs` (local docs)
7+## First Read
168  
17−## Work Management
18−This project tracks work with `bw` (Beadwork), which persists plans, progress,
19−and decisions to git so they survive compaction, session boundaries, and context
20−loss.
9+1. Read `.spec/README.md`.
10+2. Read `.spec/decisions/README.md` and any ADRs that affect the subject you are changing.
11+3. Read the current `.spec/specs/*.spec.md` files before editing.
2112  
22−- Always run `bw prime` before starting implementation work.
23−- Use one Beadwork issue per branch/PR unless the issue is explicitly split.
24−- Start work with `bw start <id>` and record validation, skip decisions, PR
25− links, and CI outcomes with `bw comment <id> "..."`
26−- Completing a task includes committing, pushing, closing the Beadwork issue,
27− and running `bw sync`.
28−- For architecture-review findings, prefer the portable helper:
29− `scripts/bw-codex-loop.sh`.
13+## Working Rules
3014  
31−## Architecture Snapshot
32−- `Jido.Agent`: pure agent module with immutable state and `cmd/2`
33−- `Jido.AgentServer`: GenServer runtime for directives, lifecycle, and message flow
34−- `Jido.Agent.Directive.*`: typed effect descriptors (`Emit`, `SpawnAgent`, `StopChild`, etc.)
35−- `Jido.Agent.StateOp.*`: internal state transition operations applied by strategy layer
36−- Plugins/sensors provide capability composition without coupling core agent logic
37− 
38−## Standards
39−- Keep `cmd/2` pure: same input => same `{agent, directives}` output
40−- Keep directives for external effects only; do not rely on directives for state mutation
41−- Use **Zoi-first** schemas for new agent/plugin/signal contracts
42−- Preserve tagged tuple and structured error contracts at public boundaries
43−- Keep cross-agent communication on signals/directives, not ad-hoc process messages
44− 
45−## Testing and QA
46−- Prefer pure agent tests first, then AgentServer/runtime integration tests
47−- Use helpers from `test/AGENTS.md` (`JidoTest.Case`, `JidoTest.Eventually`) for async assertions
48−- Avoid `Process.sleep/1` in tests; assert eventual state/event behavior
49− 
50−## Release Hygiene
51−- Keep semver ranges stable (`~> 2.0` for Jido ecosystem peers)
52−- Use Conventional Commits
53−- Do not modify `CHANGELOG.md`; release notes are generated from Git history during release, so keep changes focused on proper Conventional Commits.
54− 
55−## References
56−- `README.md`
57−- `usage-rules.md`
58−- `guides/`
59−- `test/AGENTS.md`
60−- https://hexdocs.pm/jido
15+- Keep one subject per file.
16+- Put normative statements in `spec-requirements`.
17+- Add `spec-scenarios` only when `given` / `when` / `then` improves clarity.
18+- Add `spec-meta.decisions` only when a subject depends on a durable cross-cutting ADR.
19+- Keep ADRs in `.spec/decisions/*.md` for cross-cutting policy only.
20+- Prefer targeted command verifications for behavioral proof.
21+- Use file-backed verifications only when the target can carry stable `covers:` markers for every covered id.
22+- Keep verification targets repository-root-relative.
23+- Use Git history and pull requests as the change log; keep `.spec` current-state only.
24+- At session start, run `mix spec.prime --base HEAD`.
25+- After code, docs, or tests change, run `mix spec.next`.
26+- For bug fixes, prefer `mix spec.next --bugfix`.
27+- If next says `needs subject updates`, update the named subject before you finish.
28+- If next says `ready for check`, move to `mix spec.check --base ...`.
29+- Use `mix spec.validate --debug` when you need low-level verifier output.
30+- Run `mix spec.status` when you need coverage or weak-spot summaries.
6131  

Also from Kynth Studios

Built for the same person as RuleStack

ToolDrift

What the AI coding tools changed last night

tooldrift.kynth.studio

StillShipping

Which agent tools have stopped shipping

stillshipping.kynth.studio

BlockDex

Search inside every shadcn registry

blockdex.kynth.studio

The studio list

One product, taken apart, once a month

Kynth Studios pulls one shipped product open every month — what it does, what it cost to build, what the pipeline behind it looks like, and what the numbers did. One email a month, nothing in between.

Double opt-in — we send one confirmation link and nothing else until you click it.

RuleStack

Built by

Kynth Studios

the studio behind ToolDrift, StillShipping and BlockDex

part of Toolproof, the measurement layer for AI agent tooling

Directory

Configs
Stacks
Compare formats
AGENTS.md vs CLAUDE.md
Cursor rules alternatives
Diff two configs
Best AGENTS.md examples
Best Cursor rules examples
What goes in a CLAUDE.md

Formats

AGENTS.md
CLAUDE.md
Cursor rules
Copilot instructions

Reference

Read API
Corpus health
Privacy Policy
Terms

© 2026 RuleStack. A Kynth Studios product. Changelog

RuleStack

The studio list

One product, taken apart, once a month

Kynth Studios pulls one shipped product open every month — what it does, what it cost to build, what the pipeline behind it looks like, and what the numbers did. One email a month, nothing in between.

Double opt-in — we send one confirmation link and nothing else until you click it.

RuleStack

Built by

Kynth Studios

the studio behind ToolDrift, StillShipping and BlockDex

part of Toolproof, the measurement layer for AI agent tooling

Directory

Configs
Stacks
Compare formats
AGENTS.md vs CLAUDE.md
Cursor rules alternatives
Diff two configs
Best AGENTS.md examples
Best Cursor rules examples
What goes in a CLAUDE.md

Formats

AGENTS.md
CLAUDE.md
Cursor rules
Copilot instructions

Reference

Read API
Corpus health
Privacy Policy
Terms

© 2026 RuleStack. A Kynth Studios product. Changelog

RuleStack

The studio list

One product, taken apart, once a month

Kynth Studios pulls one shipped product open every month — what it does, what it cost to build, what the pipeline behind it looks like, and what the numbers did. One email a month, nothing in between.

Double opt-in — we send one confirmation link and nothing else until you click it.

RuleStack

Built by

Kynth Studios

the studio behind ToolDrift, StillShipping and BlockDex

part of Toolproof, the measurement layer for AI agent tooling

Directory

Configs
Stacks
Compare formats
AGENTS.md vs CLAUDE.md
Cursor rules alternatives
Diff two configs
Best AGENTS.md examples
Best Cursor rules examples
What goes in a CLAUDE.md

Formats

AGENTS.md
CLAUDE.md
Cursor rules
Copilot instructions

Reference

Read API
Corpus health
Privacy Policy
Terms

© 2026 RuleStack. A Kynth Studios product. Changelog

RuleStack