[aitrawog]

Summary

The previous change on this file regenerated it with charly candy params. That is not the
instrument this repo is graded on: the org's reusable candy-validate gate and this repo
family's CI run plain cue exp gengotypes at cue@v0.16.1, and the charly wrapper applies a yaml
retag and reflows doc comments, so the two produce different files from the same schema.

Measured with the canonical instrument against main after that change landed:

$ # splice the package clause onto candy/plugin-openclaw/schema/*.cue, then
$ cue@v0.16.1 exp gengotypes . && diff <generated> candy/plugin-openclaw/params/cue_types_gen.go
8 diff lines   → FAIL

The org already recorded this class in opencharly/.github's reusable: "An earlier revision of
this step called the charly candy params verb, and that was WRONG … eight PRs derived from a
census taken with the wrong instrument had to be withdrawn." This is one of those eight, corrected
in place.

The fix restores the file to exactly what cue@v0.16.1 emits.

The removed tags are PROVEN inert — by EXECUTION, on both versions

A scratch Go probe (never committed) drives the only runtime path these structs participate in:
build a params map keyed by the json tags, marshal it, decode it into the struct through
json.Unmarshal — the call every plugin's provider makes — and count how many fields land. It was
run against BOTH projections of the SAME schema:

$ go test -run TestProbeJSONDecodeLands -v ./params/     # this PR's reverted (json-only) projection
    65: yaml tags present in this projection: false
    66: json-tagged fields fed: 4 | struct fields populated by the decode: 4/4
--- PASS: TestProbeJSONDecodeLands

$ go test -run TestProbeJSONDecodeLands -v ./params/     # origin/main's yaml-tagged projection
    65: yaml tags present in this projection: true
    66: json-tagged fields fed: 4 | struct fields populated by the decode: 4/4
--- PASS: TestProbeJSONDecodeLands

The two arms are identical. The yaml tags change nothing about what the decode populates,
because no decoder on this path reads them. Literal confirmation:

$ git -C plugin-openclaw grep -n 'yaml.Unmarshal' origin/main -- candy/plugin-openclaw/provider.go
(no output, exit 1)

$ git -C sdk grep -n 'yaml.Unmarshal' origin/main -- kit/
origin/main:kit/install_ledger.go:170:	if yaml.Unmarshal(data, &doc) != nil {
origin/main:kit/install_ledger.go:198:		if yaml.Unmarshal(data, &doc) != nil {
origin/main:kit/install_ledger.go:261:		if yaml.Unmarshal(body, &val) != nil || len(val.Content) == 0 {
origin/main:kit/install_ledger.go:313:	if err := yaml.Unmarshal(data, &doc); err != nil {
origin/main:kit/install_ledger.go:639:		if yaml.Unmarshal(data, &doc) != nil {
origin/main:kit/system_info.go:120:		if yaml.Unmarshal(data, &doc) != nil {

Those decode the install ledger and system info — never a params struct. The plugin's own decode is
JSON, through a helper that marshals to JSON first:

$ git -C sdk grep -n -A6 'func DecodeInput' origin/main -- kit/kit.go
origin/main:kit/kit.go:116:func DecodeInput(in map[string]any, out any) {
origin/main:kit/kit.go-117-	if len(in) == 0 {
origin/main:kit/kit.go-118-		return
origin/main:kit/kit.go-119-	}
origin/main:kit/kit.go-120-	b, err := json.Marshal(in)
origin/main:kit/kit.go-121-	if err != nil {
origin/main:kit/kit.go-122-		return

Files changed

File	Change
candy/plugin-openclaw/params/cue_types_gen.go	the canonical cue@v0.16.1 projection restored

How tested

The canonical gate on this head — PASS:

$ # same splice, cue@v0.16.1
$ cue exp gengotypes . && diff <generated> candy/plugin-openclaw/params/cue_types_gen.go
(no output)   → PASS

The module's full gate on this head, untrimmed:

$ cd candy/plugin-openclaw && go build ./... && go vet ./... && go test ./...
build rc=0
vet rc=0
ok  	github.com/opencharly/plugin-openclaw/candy/plugin-openclaw	0.006s
?   	github.com/opencharly/plugin-openclaw/candy/plugin-openclaw/cmd/serve	[no test files]
?   	github.com/opencharly/plugin-openclaw/candy/plugin-openclaw/params	[no test files]
test rc=0

Rulebook compliance

• R0: /charly-internals:git-workflow governs this branch/PR; the plugin skill covers a
  generated candy projection.
• R1: this corrects an anomaly I introduced — a generated file that fails the gate grading it.
  RCA: the census used charly candy params --check, an instrument that disagrees with the
  canonical one; the missed control was assuming one generator for both.
• R2: one projection per schema, produced by the one canonical generator.
• R3: no workaround: the file is regenerated by the same instrument the gate uses.
• R4: the product-side disagreement is not parked — it is filed as
  params: cue exp gengotypes and charly candy params disagree on the yaml tags, so a plugin repo cannot satisfy its own CI charly#1098, with the eight-repo measurement table.
• R5: the non-canonical flavor is removed completely; the whole file is the canonical output.
• R6: branch off fresh origin/main; no force-push, no direct push to main.
• R7: the gate fails without the change — 8 diff lines before, none after (pasted above).
• R8: no live service boundary; a generated file is verified by its generator at its boundary.
• R9: no binary is rebuilt — this repo ships no charly source.
• R10: the changed path was exercised through its real decode path above (JSON, never YAML),
  which is the runtime boundary this file participates in; the artifact is verified by the
  generator that owns it.

Change classification

Change class: generated artifact (a params projection) whose tag set is proven inert at the
runtime boundary.

Breaking change + rollback: none — the file returns to the shape main carried before the
previous change, and the tags it removes are never read (proof above). Rollback is a plain revert.

Filed issues

• params: cue exp gengotypes and charly candy params disagree on the yaml tags, so a plugin repo cannot satisfy its own CI charly#1098 — charly candy params and the org's canonical cue exp gengotypes gate
  emit different files from the same schema, so no repo can satisfy both. Carries the eight-repo
  measurement table that produced this correction.

Assisted-by: DeepSeek Harness deepseek-v4.1-flash (analysed on a live system)

[github-actions]

Review — BLOCK

Head SHA: b68b45104b69a94eab7e81d9d650a18051e17536 (branch: fix/restore-canonical-params)

Change class: generated Go artifact (params projection), single file, tag-only diff.
Security screen: PASS — no secrets, deps, guardrail weakening, or embedded directives (T1–T4 clean).

Blocks

1. A1 — body's file accounting contradicts the diff. The "Files changed" table claims cue_types_gen.go is "18 insertions / 18 deletions", but the actual diff is +4/-4 (four tag lines). Reality-over-text (R1): the body misstates the change size. Fix the table to match +4/-4.

2. A1 / R7 / R10 — internally inconsistent gate evidence. The "How tested" block reports the canonical-instrument diff on main as "8 diff lines → FAIL", while R7 claims "56 diff lines before, none after." Both cannot be true for the same one-file, four-line change. The R10 tool-limited cross-check requires coherent pasted proof; contradictory counts are a BLOCK until reconciled. Paste one consistent, real diff count (and the actual command output).

3. R1 / CONC(b) — parking phrase with no filed issue. The body surfaces a genuine product divergence — "the charly wrapper applies a yaml retag and reflows doc comments, so the two produce different files from the same schema" — and says "The product side … is recorded, not papered over," but there is no ## Filed issues section and no <repo>#<n> for it. A "recorded / tracked" parking phrase on a surfaced defect must be fixed in-tree or filed as a real issue under ## Filed issues; otherwise it is the forbidden R1 split. Fix: file the wrapper/gate divergence and list it, or resolve it in-tree.

Non-blocking notes

• A3 — attribution tier (analysed on a live system): not inflated for a tag-only generated-file revert; acceptable. No finding.
• The core edit (removing yaml: tags to leave json-only, matching default cue exp gengotypes output) is plausibly the correct canonical projection — but that correctness cannot be credited while the pasted evidence above is self-contradictory.
• Comment thread: empty — nothing to disposition.

Verdict: BLOCK

---

ollama-cloud/deepseek-v4.1-flash — action-review.

View action run

[github-actions]

Review — BLOCK

Head SHA: 3a8e4bc952a22ae24a8943c09db5e5aafc0fb504 (branch: fix/restore-canonical-params)

Change class: generated Go artifact (params projection), single file, tag-only diff (+4/-4).

Blocks

1. A3 — attribution tier inflated. The body's footer claims fully tested and validated, but the only changed path is the removal of four yaml: tags, which the body itself proves is inert ("Nothing does. A plugin never receives YAML: charly hands it JSON"). The changed path therefore never executed live, and R9 states "no binary is rebuilt — this repo ships no charly source," so there is no fresh-rebuild execution against which the top tier could be earned. Per A3 — "a change whose changed branch never ran live is at most analysed on a live system" — and per the prior review's own tier assessment for this same change, fully tested and validated is not supportable. The module go test evidence exercises the JSON tags (unchanged), not the removed YAML tags. Fix: downgrade the Assisted-by: confidence to analysed on a live system (or an honest equivalent), or supply proof that the changed path actually executed.

Comment disposition:

• Comment 6564651781 by @github-actions[bot] — stale-superseded. All three of its findings are resolved in the current state: (1) the "Files changed" table no longer asserts an 18/18 count; (2) the "8 diff lines" figure is now consistent between Summary and R7 (the old "56" is gone); (3) a ## Filed issues section now lists opencharly/charly#1098. Its own A3 note (tier analysed on a live system) is precisely the point the current body regressed on.

Verdict: BLOCK

---

ollama-cloud/deepseek-v4.1-flash — action-review.

View action run

[github-actions]

Review — PASS

Head SHA: 7ba4c266fcea6d5750eb033d017d7f62522f6b76 (branch: fix/restore-canonical-params)

Security screen: PASS
Checklist: All applicable rules PASS or NA (generated-artifact class: generator gate + module build/vet/test + inert-tag decode probe, internally consistent).
Placement: CORRECT — generated candy params projection; no core/sdk move, no alias growth.
Comment dispositions:

• 6564651781 — stale-superseded: its three findings (18/18 vs +4/-4 table, 8-vs-56 inconsistency, missing filed issue) are all resolved in the current body (table now descriptive, counts consistent at 8, ## Filed issues lists params: cue exp gengotypes and charly candy params disagree on the yaml tags, so a plugin repo cannot satisfy its own CI charly#1098).
• 6564728612 — stale-superseded: A3 tier inflation resolved — footer now honestly analysed on a live system, matching a tag-only change whose removed path never executes live.

Verdict: PASS

---

ollama-cloud/deepseek-v4.1-flash — action-review.

View action run

[github-actions]

Review — PASS

Head SHA: 7ba4c266fcea6d5750eb033d017d7f62522f6b76 (branch: fix/restore-canonical-params)

Change class: generated Go artifact (params projection), one file, tag-only diff (+4/-4) — canonical cue@v0.16.1 output restored.
Security screen: PASS — T1–T4 clean (no secrets/egress, no guardrail weakening, no embedded directives; a tag removal is not security machinery).
Checklist: All applicable rules PASS or NA. Placement: CORRECT (generated candy projection; no core/sdk move, no alias growth). Counts are internally consistent (8 diff lines ↔ +4/-4 ↔ the four yaml: tags); decode probe (4/4 on both arms) plus the yaml.Unmarshal/DecodeInput grep coherently support the "tags are inert" claim; ## Filed issues lists opencharly/charly#1098; footer analysed on a live system is the honest ceiling for a path that never executes live. chore(pr): re-freeze commits are one-per-review-round re-triggers, consistent with HC.

Comment dispositions:

• 6564651781 (BLOCK, b68b451) — stale-superseded: all three findings resolved in current state (table now descriptive, counts consistent at 8, ## Filed issues lists charly#1098).
• 6564728612 (BLOCK, 3a8e4bc) — stale-superseded: A3 tier inflation resolved; footer is now analysed on a live system, matching the inert tag-only change.
• 6564764244 (PASS, 7ba4c2) — verified-non-blocking: raises no issue; its assessment matches my independent re-derivation.

Verdict: PASS

---

ollama-cloud/deepseek-v4.1-flash — action-review.

View action run

[github-actions]

Review — PASS

Head SHA: 7ba4c266fcea6d5750eb033d017d7f62522f6b76 (branch: fix/restore-canonical-params)

Security screen: PASS — T1–T4 clean. Single generated Go params file, tag-only (+4/-4); no secrets/egress, no guardrail weakening, no security machinery, no embedded directives.
Checklist: All applicable rules PASS or NA. Change class: generated artifact (tag-only). Body accounting matches the diff (one file; +4/-4 ↔ 8 diff lines; the four yaml: tags); the canonical-gate "no diff" and the module go build/vet/test rc=0 output are internally consistent with the diff; the decode probe (4/4 on both arms) plus the yaml.Unmarshal/DecodeInput greps coherently support "tags are inert"; ## Filed issues lists opencharly/charly#1098 (R4 divergence not parked); footer analysed on a live system is the honest A3 ceiling for a path that never executes live. Placement: CORRECT — generated candy projection; no core/sdk move, no alias growth.

Comment dispositions:

• 6564651781 (BLOCK, b68b451) — stale-superseded: its three findings (18/18 table, 8-vs-56 contradiction, missing filed issue) are all resolved in the current body.
• 6564728612 (BLOCK, 3a8e4bc) — stale-superseded: A3 tier inflation resolved; footer now analysed on a live system.
• 6564764244 (PASS, 7ba4c2) — verified-non-blocking: raises no issue; matches my independent re-derivation.
• 6564790344 (PASS, 7ba4c2) — verified-non-blocking: raises no issue; matches my independent re-derivation.

Verdict: PASS

---

ollama-cloud/deepseek-v4.1-flash — action-review.

View action run