@lmzhen/dsh-tool-skill-manage 0.3.83 → 0.4.1

package/README.md

@@ -1,36 +1,31 @@
 # @lmzhen/dsh-tool-skill-manage
 
-Model-facing skill_manage tool
+Model-facing `skill_manage` tool: it exposes skill-library mutations to the model and
+routes them through the evolution approval seam when one is mounted; it owns neither the
+library nor its lifecycle rules.
 
-## Model Experience
+## Model surface
 
-### skill_manage tool surface
-
-#### What the model sees
-
-The model sees the `skill_manage` tool schema and tool results containing success or validation messages.
-
-#### Token effect
-
-Adds the `skill_manage` tool schema to the request catalog; result tokens scale with returned messages.
-
-#### KV Cache effect
-
-Tool schema is prefix-stable. Skill writes do not alter the current request prompt; catalog invalidation affects the next request.
+- **Model-visible:** the `skill_manage` tool schema and its success/validation messages; result tokens scale with what is returned.
+- **Prompt prefix / KV cache:** the tool schema is prefix-stable, and skill writes do not alter the current request prompt; catalog invalidation affects the next request; family rules: `packages/README.md` §"Model-visible prompt prefix and the KV cache".
+- **Mount it?** yes — the `tool-skill-manage` row, carried by the `evolution-all` and one-click `evolution-preset` bundles and the Evolution agent preset delta.
 
 ## Safety model
 
 ### Approval seam
 
-Mutations (create/edit/update/patch/delete/write_file/remove_file/restructure) pass through the evolution approval seam when `evolution-approval` is mounted; approved/staged writes are replayed by the registered runner with the library origin preserved.
-
-### pin/unpin: explicit exception (0.3.18, E-70)
+Mutations (create/edit/update/patch/delete/write_file/remove_file/restructure) pass through the evolution approval seam when `evolution-approval` is mounted; approved/staged writes are replayed by the registered runner with the library origin preserved. The missing-required-argument pre-check runs BEFORE that boundary — one shared table and one refusal builder with `executeCore`, so the stage boundary and the execution check cannot diverge: a write that cannot execute is refused, never staged for approval.
 
-`pin` and `unpin` are deliberately **outside** the approval seam. Pinning only lifts/restores the curator-lifecycle freeze — it changes a lifecycle flag, never content — and is fully reversible by the same tool. Routing it through `policy:'ask'` would let a staged-but-never-approved request hold the library in a pinned state invisibly. Tradeoff accepted: no approval on a lifecycle-flag flip; if product policy changes, pin/unpin should be wired into the same staging path as `patch`.
+### pin/unpin: explicit exception
 
-## Known Limitations and Deferred Work
+`pin` and `unpin` are deliberately **outside** the approval seam. Pinning only lifts/restores the curator-lifecycle freeze (a lifecycle flag, never content) and is fully reversible by the same tool. Routing it through `policy:'ask'` would let a staged-but-never-approved request hold the library in a pinned state invisibly. Tradeoff accepted: no approval on a lifecycle-flag flip; if product policy changes, pin/unpin should be wired into the same staging path as `patch`.
 
+## Known limitations
 
 - No known durable consumer gaps at this time. Runtime contracts are covered by package and boundary tests.
 
 **Runtime invariant:** No companion is published. The platform auto-assembles nothing and the family mounts no `<pkg>/invariant` cordis row, so a companion here would never execute (v37 S2.1 / I-3).
+
+## Notes and history
+
+- pin/unpin: explicit exception (0.3.18, E-70)

package/lib/index.js

@@ -8,8 +8,11 @@ import { DEFAULT_SKILL_LIMITS, DSH_AUTHORING_STANDARDS, SKILLS_GUIDANCE, SKILLS_
 *
 * Mutations pass through the evolution approval seam when it is mounted;
 * approved/staged background writes are replayed by the registered runner.
-* The skill library itself never hard-deletes: archive is the maximum
-* destructive action and every curator mutation is snapshot-reversible.
+* On LIVE skills, archive is the maximum destructive action and every
+* curator mutation is snapshot-reversible. Two retention hard-deletes exist
+* by design, both mtime-gated and logged: expired ARCHIVES are pruned after
+* 365 days (skill-store `pruneExpiredArchives`), and pre-run snapshots are
+* capped at the newest 5 — restore-ability holds inside those windows.
 *
 * 0.3.18 (E-70) — pin/unpin are deliberately OUTSIDE the approval seam:
 * pinning only lifts/restores the curator-lifecycle freeze and is fully
@@ -358,15 +361,15 @@ function apply(ctx, rawConfig = {}) {
 			const approval = ctx.get("evolutionApproval");
 			if (approval && args.action !== "list" && args.action !== "review" && args.action !== "pin" && args.action !== "unpin") {
 				if ((args.action === "update" || args.action === "edit") && typeof args.name === "string" && args.name !== "") {
-					const stageCurrent = await library.read(args.name).catch(() => void 0);
-					if (stageCurrent !== void 0) operation = {
+					const stageCurrent = await library.read(args.name).catch(() => null);
+					operation = {
 						...args,
 						staged_from_sha256: stageCurrent === null ? "absent" : contentHash(stageCurrent)
 					};
 				}
 				if ((args.action === "write_file" || args.action === "remove_file") && typeof args.name === "string" && args.name !== "" && typeof args.file_path === "string" && args.file_path !== "") {
-					const stageFile = await library.readSupportFile(args.name, args.file_path).catch(() => void 0);
-					if (stageFile !== void 0) operation = {
+					const stageFile = await library.readSupportFile(args.name, args.file_path).catch(() => null);
+					operation = {
 						...args,
 						staged_from_sha256: stageFile === null ? "absent" : contentHash(stageFile)
 					};

package/lib/types/index.d.ts

@@ -3,8 +3,11 @@
  *
  * Mutations pass through the evolution approval seam when it is mounted;
  * approved/staged background writes are replayed by the registered runner.
- * The skill library itself never hard-deletes: archive is the maximum
- * destructive action and every curator mutation is snapshot-reversible.
+ * On LIVE skills, archive is the maximum destructive action and every
+ * curator mutation is snapshot-reversible. Two retention hard-deletes exist
+ * by design, both mtime-gated and logged: expired ARCHIVES are pruned after
+ * 365 days (skill-store `pruneExpiredArchives`), and pre-run snapshots are
+ * capped at the newest 5 — restore-ability holds inside those windows.
  *
  * 0.3.18 (E-70) — pin/unpin are deliberately OUTSIDE the approval seam:
  * pinning only lifts/restores the curator-lifecycle freeze and is fully

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@lmzhen/dsh-tool-skill-manage",
   "description": "Model-facing skill_manage tool (community build)",
-  "version": "0.3.83",
+  "version": "0.4.1",
   "publishConfig": {
     "access": "public"
   },
@@ -27,24 +27,24 @@
   "license": "MIT",
   "dependencies": {
     "@deepseek-ai/schemastery": "^3.18.1",
-    "@lmzhen/dsh-evolution-approval": "^0.3.83",
-    "@lmzhen/dsh-evolution-core": "^0.3.83"
+    "@lmzhen/dsh-evolution-approval": "^0.4.1",
+    "@lmzhen/dsh-evolution-core": "^0.4.1"
   },
   "peerDependencies": {
     "@deepseek-ai/cordis": "^4.0.1",
     "@deepseek-ai/dsh-skill": "^0.1.5-rc.2",
     "@deepseek-ai/dsh-system-prompt": "^0.1.5-rc.2",
     "@deepseek-ai/dsh-tools": "^0.1.5-rc.2",
-    "@lmzhen/dsh-evolution-io": "^0.3.83",
-    "@lmzhen/dsh-skill-usage": "^0.3.83"
+    "@lmzhen/dsh-evolution-io": "^0.4.1",
+    "@lmzhen/dsh-skill-usage": "^0.4.1"
   },
   "devDependencies": {
     "@deepseek-ai/dsh-agent-loop-testkit": "^0.1.5-rc.2",
     "@deepseek-ai/dsh-skill": "^0.1.5-rc.2",
     "@deepseek-ai/dsh-system-prompt": "^0.1.5-rc.2",
     "@deepseek-ai/dsh-tools": "^0.1.5-rc.2",
-    "@lmzhen/dsh-evolution-core": "^0.3.83",
-    "@lmzhen/dsh-evolution-io": "^0.3.83",
-    "@lmzhen/dsh-skill-usage": "^0.3.83"
+    "@lmzhen/dsh-evolution-core": "^0.4.1",
+    "@lmzhen/dsh-evolution-io": "^0.4.1",
+    "@lmzhen/dsh-skill-usage": "^0.4.1"
   }
 }