Prove durable Hello Echo execution

[flyingrobots]

Runs the compiler-produced Hello Echo package through Echo’s generic scheduler-owned runtime and proves WAL ordering, recovery, typed duplicate obstruction, altered-basis refusal, and replay equivalence.

Closes #3.

[flyingrobots]

@coderabbitai review

[flyingrobots]

@codex please review

[coderabbitai]

✅ Action performed

Review finished.

Note: CodeRabbit is an incremental review system and does not re-review already reviewed commits. This command is applicable only when automatic reviews are paused.

[coderabbitai]

Summary by CodeRabbit

• New Features

  • Added a local runtime witness suite covering compilation, durable WAL handling, recovery, duplicate operations, deterministic reruns, malformed inputs, byte limits, and bounded stress scenarios.
  • Added a command-line test runner for executing runtime operations with JSON inputs and isolated WAL directories.
  • Added a greeting input fixture for runtime validation.

• Documentation

  • Documented runtime witness setup, supported scenarios, verification guarantees, and roadmap status.

Walkthrough

Adds a shell-based runtime witness that invokes compiler-produced artifacts through Echo, validates durable recovery and deterministic replay, exercises input boundaries and stress cases, and documents the singleton scheduler scope.

Changes

Runtime witness

Layer / File(s)	Summary
Execution harness tests/create-greeting.json, tests/run.sh, tests/runtime.sh	Creates isolated runtime cases, validates required artifacts and environment variables, and invokes the Edict operation runner with JSON input and WAL paths.
Witness validation matrix tests/runtime.sh	Validates witness structure, golden execution, byte-identical replay, path portability, generated artifacts, malformed input, 256/257-byte boundaries, fixed-seed typed cases, and eight recovery stress cases.
Documented witness contract README.md, docs/roadmap.md, CHANGELOG.md	Documents the runtime command, proven durable behaviors, singleton scheduler boundary, and roadmap status.

Estimated code review effort: 4 (Complex) | ~45 minutes

Sequence Diagram(s)

sequenceDiagram
  participant Developer
  participant runtime.sh
  participant run.sh
  participant Echo
  participant jq
  Developer->>runtime.sh: run runtime witness suite
  runtime.sh->>run.sh: provide typed input and isolated WAL path
  run.sh->>Echo: invoke run-edict-operation with compiler artifacts
  Echo-->>run.sh: return witness JSON or validation error
  run.sh-->>runtime.sh: write witness.json
  runtime.sh->>jq: validate execution, recovery, replay, and boundary results

Loading

Possibly related issues

• flyingrobots/echo issue 690 — Covers the scheduler-owned, WAL-backed Hello Echo runtime witness exercised here.
• flyingrobots/echo issue 699 — Covers duplicate no-mutation evidence validated by the runtime witness.

Poem

WAL before the welcome bell,
Echo wakes and guards the cell.
Replay leaves each byte in line,
Eight small worlds recover fine.
One Tick proves the path is sound.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	Clearly names the main change: proving durable Hello Echo execution.
Description check	✅ Passed	It directly describes the new durable runtime witness and recovery checks.
Linked Issues check	✅ Passed	The runtime suite and scripts cover the scheduler-owned witness, recovery, duplicate handling, altered-basis refusal, and path hygiene.
Out of Scope Changes check	✅ Passed	The changes stay within the runtime witness, supporting scripts, docs, and one test fixture; no unrelated code paths were added.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@tests/runtime.sh`:
- Around line 72-75: Update the disclosure scan around the golden_witness grep
to canonicalize both EDICT_REPO and ECHO_REPO using cd ... && pwd -P before
building the grep patterns. Use the resolved absolute paths, rather than raw
repository values, so relative checkouts match leaked canonical paths and "."
does not create broad false-positive patterns.
- Around line 26-54: Update assert_common_witness in tests/runtime.sh to
independently compute SHA-256 digests for the package, verification report, and
lawpack manifest files supplied to run-edict-operation, then assert those
digests equal the pinned constants before validating the witness claims. Reuse
the actual input-file variables or arguments from the surrounding runner flow,
and preserve the existing witness-field checks while requiring both independent
hashes and reported artifact digests to match.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 23a0e575-86eb-47f0-8fb1-3e145ea923f0

📥 Commits

Reviewing files that changed from the base of the PR and between 8dd96de and a8e2b79.

📒 Files selected for processing (6)

• CHANGELOG.md
• README.md
• docs/roadmap.md
• tests/create-greeting.json
• tests/run.sh
• tests/runtime.sh

📜 Review details

🧰 Additional context used

🪛 Shellcheck (0.11.0)

tests/run.sh

[warning] 9-9: Remove space after = if trying to assign a value (for empty string, use var='' ... ).

(SC1007)

🔇 Additional comments (5)

tests/create-greeting.json (1)

1-5: LGTM!

tests/runtime.sh (1)

9-24: LGTM!

Also applies to: 57-70, 77-175

tests/run.sh (1)

1-49: LGTM!

docs/roadmap.md (1)

77-80: LGTM!

CHANGELOG.md (1)

15-18: LGTM!

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: a8e2b79bf5

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[flyingrobots]

@coderabbitai review

[flyingrobots]

@codex please review

[coderabbitai]

✅ Action performed

Review finished.

Note: CodeRabbit is an incremental review system and does not re-review already reviewed commits. This command is applicable only when automatic reviews are paused.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: d55a2e8f46

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[flyingrobots]

AUTHORIZED — one additional Code Lawyer remediation pass for PR #17.

This is an operator-approved exception after the configured two-pass budget
performed its intended escalation function. It does not reset or remove the
budget.

Authorized finding:
PRRT_kwDOTmLBK86UkvoG — P1 “Verify duplicate execution leaves state unchanged”

Required outcome:

• Expose generic, independently comparable application-state evidence before
  and after the duplicate attempt.
• The evidence must be produced by Echo’s generic runner, not reconstructed
  using Hello Echo application semantics.
• If the producer change belongs in Echo, route it through a separate Echo
  issue and PR before updating Hello Echo PR Prove durable Hello Echo execution #17.
• Update tests/runtime.sh to require both roots or equivalent authoritative
  digests to be present, valid, and equal.
• Preserve the existing internal fail-closed hidden-mutation check.
• Add or retain a negative runner test proving that a duplicate path which
  mutates authoritative application state cannot produce a passing witness.
• Reply to and resolve the named bot-authored thread after validation.

Scope:

• Only changes necessary to remediate the named finding.
• No opportunistic refactoring.
• No application-specific behavior in Echo.
• No native Hello Echo callback or handwritten result reconstruction.
• No weakening of existing tests, gates, or documentation claims.

Validation:

• Relevant Echo runner and report-schema tests.
• Echo formatting and clippy for affected crates.
• Full Hello Echo runtime witness suite.
• ShellCheck, Markdown lint, and git diff --check.
• Clean worktrees and published feature branches.

Closure:

• Re-evaluate the merge gate after the remediation.
• Merge PR Prove durable Hello Echo execution #17 only when the named thread is resolved, all required validation
  passes, and no unresolved bot-authored review thread remains.
• No fourth Code Lawyer pass is authorized. Any new actionable thread causes a
  new escalation rather than another autonomous remediation cycle.

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@tests/runtime.sh`:
- Around line 74-85: The runtime test currently verifies only unchanged state
and digests, without exercising duplicate-mutation rejection. Add a negative
generic-runner test near the existing duplicate assertions that mutates
duplicate-path state and verifies the witness fails, then update the related
README and roadmap claims to reflect the tested guarantee.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 4f33b94d-f5fa-4459-9b65-d4449c8c2d44

📥 Commits

Reviewing files that changed from the base of the PR and between a8e2b79 and 00f6204.

📒 Files selected for processing (5)

• CHANGELOG.md
• README.md
• docs/roadmap.md
• tests/run.sh
• tests/runtime.sh

📜 Review details

🔇 Additional comments (3)

tests/runtime.sh (1)

15-31: LGTM!

Also applies to: 46-63, 107-136, 227-227

tests/run.sh (1)

6-53: LGTM!

CHANGELOG.md (1)

15-28: LGTM!

[flyingrobots]

Merge gate open at 00f6204a233a762e0d7d4df7cd0ff8ee4fbd3ad4.

• Full Hello Echo runtime matrix passed.
• ShellCheck, Markdown lint, and diff checks passed.
• CodeRabbit status is successful.
• Zero unresolved review threads remain.
• PR is mergeable with a CLEAN merge state.
• The sole new actionable thread was satisfied by producer-owned Echo #700 evidence and resolved under explicit operator authority.
• This repository is solo-maintained and main has no required-review protection; criteria above substitute for an additional human approval.