feat(code-review): add review package creation and updating - #50
colinhallett wants to merge 12 commits into
Conversation
|
@tessl-code-review Please review the package creation and updating workflows, active-profile preservation, hosted result contract, and evaluation coverage. |
There was a problem hiding this comment.
Tessl Code Review
Changes requested (1 required, 3 advisory)
The new package workflows establish local-package creation, preservation, and evidence-guided maintenance, but their dispatch and hosted terminal-result instructions leave several supported requests routed or reported incorrectly. The hosted result example also contradicts its creation contract.
| Severity | Requires changes | Advisory |
|---|---|---|
| Major | 1 | 0 |
| Minor | 0 | 3 |
Requires changes (1)
- Major · Requires changes · Package-maintenance requests are not routed to the update workflow
code-review/skills/code-review/SKILL.md:47
Advisory (3)
- Minor · Advisory in Standard mode · Creation stops on an existing default-path manifest without a terminal outcome
code-review/skills/create-code-review-package/SKILL.md:18 - Minor · Advisory in Standard mode · Update prerequisites can terminate or continue without the required blocked outcome
code-review/skills/update-code-review-package/SKILL.md:13 - Minor · Advisory in Standard mode · The creation result example violates the stated evidence-field contract
code-review/skills/code-review/references/hosted-authoring.md:22
Working with this review
Verify before acting. These findings are AI-generated and not authoritative. Check each one against the code, yourself or with an agent, before acting on it.
Answer each finding on its thread:
- Addressed: you changed the code. Say what changed.
- Refuted: the finding is wrong about the code.
- Declined: the finding is right, and you are not changing it. Say why.
Reviews are incremental. The first review shows every valid optional and required finding. A later round checks the latest changes and earlier findings, and adds a fresh finding only when it requires changes. It never repeats one you refuted or declined. The configured request-changes severity sets the consequence baseline, and review mode sets the strength required to block. Ask for a full review to inspect the whole change and show optional findings again.
Lenses are configurable. Add your own and scope them to paths with globs. Keep the set small so findings stay distinct. See https://docs.tessl.io/tutorials/setting-up-agentic-code-review#route-reviews-with-yaml
Mention @tessl-code-review to run another review.
|
@tessl-code-review The four findings have been addressed and answered. Please review the latest package-workflow fixes. |
There was a problem hiding this comment.
Tessl Code Review
Changes requested (1 required)
1 finding requires changes before approval.
| Severity | Requires changes | Advisory |
|---|---|---|
| Major | 1 | 0 |
Requires changes (1)
- Major · Requires changes · Interactive updates can edit a package without usable review evidence
code-review/skills/update-code-review-package/SKILL.md:21
Earlier findings: 4 addressed
About this review
These findings are AI-generated and not authoritative. Check each one against the code before acting on it. See the guidance on working with this review.
Mention @tessl-code-review to run another review.
|
@tessl-code-review The remaining interactive evidence-handling inconsistency is fixed and its thread answered. Please rereview the latest head. |
There was a problem hiding this comment.
Tessl Code Review
Changes approved
No findings to surface.
Earlier findings: 5 addressed
About this review
These findings are AI-generated and not authoritative. Check each one against the code before acting on it. See the guidance on working with this review.
Mention @tessl-code-review to run another review.
|
@tessl-code-review Please review the latest commit. The prior round approved the workflow changes; the subsequent commit corrects an evidence rubric and adds a recurring-noise regression scenario. |
There was a problem hiding this comment.
Tessl Code Review
Changes requested (1 required)
The package workflows now route active-package maintenance correctly and consistently handle blocked creation and update states. The new recurring-noise evaluation, however, permits an exclusion broader than the two evidenced generated files, which can train an update to omit security review for unrelated source in that directory.
| Severity | Requires changes | Advisory |
|---|---|---|
| Major | 1 | 0 |
Requires changes (1)
- Major · Requires changes · The recurring-noise rubric permits excluding unevidenced files under
src/generated
code-review/evals/update-review-package-recurring-noise/criteria.json:6
Earlier findings: 5 addressed
About this review
These findings are AI-generated and not authoritative. Check each one against the code before acting on it. See the guidance on working with this review.
Mention @tessl-code-review to run another review.
There was a problem hiding this comment.
Tessl Code Review
Changes requested (1 required)
The package-creation, package-update, dispatch, and hosted-result instructions now consistently cover the previously reported blocked and evidence-handling paths. The recurring-noise evaluation still allows a directory-scoped security exclusion broader than the evidence establishes.
| Severity | Requires changes | Advisory |
|---|---|---|
| Major | 1 | 0 |
Requires changes (1)
- Major · Requires changes · The recurring-noise rubric still permits excluding unevidenced files under
src/generated
code-review/evals/update-review-package-recurring-noise/criteria.json:6
Still applies. See the existing thread.
Earlier findings: 1 still applies · 5 addressed
About this review
These findings are AI-generated and not authoritative. Check each one against the code before acting on it. See the guidance on working with this review.
Mention @tessl-code-review to run another review.
|
@tessl-code-review Addressed the latest finding in d34331dcdcf4db58f7b4b52c78ec07051f8c0eed. Please review the exact-output exclusion rubric and the new unmarked-file negative case. |
There was a problem hiding this comment.
Tessl Code Review
Changes approved
1 optional suggestion. Nothing blocking.
The package workflows now route active-package maintenance, preserve blocked hosted outcomes, and constrain the recurring-noise evaluation to evidenced generated files. An interactive workflow path for an existing but unreferenced package still loops between creation and update rather than providing a usable maintenance disposition.
| Severity | Requires changes | Advisory |
|---|---|---|
| Minor | 0 | 1 |
Advisory (1)
- Minor · Advisory in Standard mode · An unreferenced destination package is routed in an interactive create/update loop
code-review/skills/create-code-review-package/SKILL.md
Still applies. Its earlier thread is resolved.
Earlier findings: 1 still applies · 5 addressed
About this review
These findings are AI-generated and not authoritative. Check each one against the code before acting on it. See the guidance on working with this review.
Mention @tessl-code-review to run another review.
|
@tessl-code-review Please review current head 751ddd6. A targeted behavioral evaluation exposed that excluding a whole generated directory could also hide future hand-authored files. The update skill now requires evidence for directory-wide generated ownership and otherwise limits exclusions to evidenced paths. Plugin version is 0.4.1; skill, brand, and diff checks pass. The targeted regression will be rerun. |
There was a problem hiding this comment.
Tessl Code Review
Changes requested (1 required, 1 advisory)
The package workflows now route active-package maintenance, preserve blocked hosted outcomes, and constrain the recurring-noise evaluation to the evidenced generated files. Interactive handling of an existing unreferenced package still loops between the two workflows, and the generated-output guidance can remove those paths from unrelated lenses through a profile-wide ignore rule.
| Severity | Requires changes | Advisory |
|---|---|---|
| Major | 1 | 0 |
| Minor | 0 | 1 |
Requires changes (1)
- Major · Requires changes · Generated-output exclusions can suppress unrelated lenses
code-review/skills/update-code-review-package/SKILL.md:33
Advisory (1)
- Minor · Advisory in Standard mode · An unreferenced destination package still sends interactive users in a create/update loop
code-review/skills/create-code-review-package/SKILL.md
Still applies at the location above. See the earlier thread.
Earlier findings: 1 still applies · 5 addressed
About this review
These findings are AI-generated and not authoritative. Check each one against the code before acting on it. See the guidance on working with this review.
Mention @tessl-code-review to run another review.
|
@tessl-code-review The exclusion-scope finding and remaining interactive redirect loop are fixed and answered in adf31a5, plugin 0.4.2. Static checks pass. The preceding targeted evaluation correctly preserved unrelated lens checks and exact path coverage; final focused behavioral checks will cover these latest instructions. Please rereview. |
There was a problem hiding this comment.
Tessl Code Review
Changes approved
No findings to surface.
Earlier findings: 7 addressed
About this review
These findings are AI-generated and not authoritative. Check each one against the code before acting on it. See the guidance on working with this review.
Mention @tessl-code-review to run another review.
Adds separate workflows to create a repository-owned review package and improve an existing one. Creation copies the default lenses, adds repository-specific coverage and configures the active profile; updating uses the current local package and review evidence while preserving manual changes.
The workflows support interactive use and unattended authoring with structured results. Six new evaluation scenarios cover creation without history, profile preservation, existing-package protection, evidence-based updates, no-change and missing state.
Validation: plugin lint, version-bump check, diff whitespace check and evaluation JSON validation passed. Live skill review and behavioral evaluations are pending; this remains a draft.