@moda-ai/cli 1.10.0 → 1.10.2

package/README.md

@@ -58,12 +58,6 @@ moda tool-failures                               # Tool failure overview
 moda tool-failure-detail <tool_name>             # Per-tool failure detail
 moda tool-failure-detail <tool_name> --include-window --window=1
 
-# Agent runs runtime (workflows = recurring class, run = one execution)
-moda workflows --time-range=7d                   # Recurring workflow catalog with health
-moda runs                                        # Last 7d of workflow executions
-moda runs --time-range=24h --outcome=partial     # Last 24h, partial completions only
-moda runs --class-id=cron:helpscout-triage-hourly # Scoped to one workflow class
-moda run <run_id>                                # One execution + its side-effect receipts
 moda loops                                       # Currently-active loop detections
 moda loop <detection_id>                         # One loop detail (signature + blast radius)
 
@@ -81,9 +75,7 @@ All commands output JSON. Pipe to `jq` for filtering:
 ```bash
 moda overview | jq '.frustrations'
 moda frustrations | jq '.frustrations[].primary_cause'
-moda runs --outcome=partial | jq '.runs[] | {job_id, intent_summary, outcome_reason}'
 moda loops | jq '.loops | sort_by(-.cycle_visits) | .[0:5] | .[].loop_signature_human'
-moda workflows | jq '.workflows | sort_by(.health) | .[] | select(.health < 0.8)'
 ```
 
 ## Prompt Management

package/dist/cli.js

@@ -2,7 +2,7 @@
 import { z } from 'zod';
 import { validateConfig, VERSION, resolveApiKey } from './config.js';
 import { callDataAPI, ApiError, getApiRequestCount, resetApiRequestCount, } from './api-client.js';
-import { OverviewSchema, ClustersSchema, ClusterConversationsSchema, ConversationsSchema, SearchSchema, WorldStateSchema, ContextSchema, AuditSchema, FrustrationsSchema, ToolFailuresSchema, ToolFailureDetailSchema, WorkflowsSchema, RunsSchema, RunDetailSchema, LoopsSchema, LoopDetailSchema, } from './schemas.js';
+import { OverviewSchema, ClustersSchema, ClusterConversationsSchema, ConversationsSchema, SearchSchema, WorldStateSchema, ContextSchema, AuditSchema, FrustrationsSchema, ToolFailuresSchema, ToolFailureDetailSchema, LoopsSchema, LoopDetailSchema, } from './schemas.js';
 import { withFrustrationAnchors, withToolFailureAnchors, embeddedWindowSize, asRecord, asString, asNumber } from './transforms.js';
 import { runPromptsCommand } from './prompts.js';
 // Coding-agent capture (shadow workspace / Shadow Spark / trace hooks) is temporarily
@@ -130,7 +130,6 @@ Commands:
   agent evidence <ref>             Expand one evidence reference
   investigate                      Rank production issues with evidence
   failures                         Show production failures worth fixing first
-  why-failed <run_id>              Explain a failed or partial production run
   ask "question"                   Ask Moda a production/harness question
   sync                             Sync the local harness graph to Moda Cloud
   sync harness                     Sync the local harness graph to Moda Cloud
@@ -149,10 +148,6 @@ Commands:
   tool-failures                    Get tool failure overview
   tool-failure-detail <tool_name>  Get per-tool failure detail
 
-  Agent runs runtime:
-  workflows                        Recurring workflow catalog (per-class 30d health)
-  runs                             List workflow executions with classified outcomes
-  run <run_id>                     Get single execution + side-effect receipts
   loops                            Currently-active loop detections
   loop <detection_id>              Single loop detail
 
@@ -216,11 +211,6 @@ World-state flags:
   --summary-only                   Return segment summaries only (omit the event stream)
   --event-limit=N                  Max world-state events to return (1-5000, default 2000)
 
-Runs flags:
-  --outcome=OUTCOME                Filter by outcome (running|succeeded|partial|failed|loop|abandoned|unknown)
-  --class-id=ID                    Filter to a single workflow class (e.g. cron:helpscout-triage-hourly)
-  --user-id=ID                     Filter to a single user
-
 Init flags:
   --yes                            Accept all defaults (non-interactive)
   --analyst=auto|claude|codex|cursor|local-scan
@@ -256,7 +246,6 @@ Examples:
   moda agent evidence doctor:harness_report_approved_synced
   moda investigate
   moda failures
-  moda why-failed 12345678-1234-1234-1234-1234567890ab
   moda ask "what should I fix first?"
   moda sync --dry-run
   moda sync harness --profile=prod
@@ -276,10 +265,6 @@ Examples:
   moda context <conversation_id> --window=3
   moda audit <conversation_id|trace_id>
   moda audit <trace_id> --kind=trace --json
-  moda workflows --time-range=7d
-  moda runs --time-range=24h --outcome=partial
-  moda runs --class-id=cron:helpscout-triage-hourly --limit=20
-  moda run 12345678-1234-1234-1234-1234567890ab
   moda loops --limit=10
   moda loop 12345678-1234-1234-1234-1234567890ab
   moda prompts status
@@ -405,7 +390,6 @@ export async function runCommand(command, positional, flags, positionals = posit
         }
         case 'investigate':
         case 'failures':
-        case 'why-failed':
         case 'ask': {
             return (await runProductionCommand(context)).exitCode;
         }
@@ -594,51 +578,6 @@ export async function runCommand(command, positional, flags, positionals = posit
             await printWithOptionalWindows(withToolFailureAnchors(data), 'examples', params.include_window === true, params.window ?? 1, context.output);
             break;
         }
-        // ─── Workflows + Runs ──────────────────────────────────────
-        case 'workflows': {
-            const args = flagsToArgs(flags);
-            const params = WorkflowsSchema.parse(args);
-            const query = new URLSearchParams();
-            if (params.time_range)
-                query.set('time_range', params.time_range);
-            if (params.user_id)
-                query.set('user_id', params.user_id);
-            const queryString = query.toString() ? `?${query.toString()}` : '';
-            const data = await callDataAPI(`/workflows${queryString}`);
-            context.output.writeData(data);
-            break;
-        }
-        case 'runs': {
-            const args = flagsToArgs(flags);
-            const params = RunsSchema.parse(args);
-            const query = new URLSearchParams();
-            if (params.time_range)
-                query.set('time_range', params.time_range);
-            if (params.outcome)
-                query.set('outcome', params.outcome);
-            if (params.class_id)
-                query.set('class_id', params.class_id);
-            if (params.user_id)
-                query.set('user_id', params.user_id);
-            if (params.limit)
-                query.set('limit', params.limit.toString());
-            if (params.offset !== undefined)
-                query.set('offset', params.offset.toString());
-            const queryString = query.toString() ? `?${query.toString()}` : '';
-            const data = await callDataAPI(`/runs${queryString}`);
-            context.output.writeData(data);
-            break;
-        }
-        case 'run': {
-            if (!positional) {
-                throw new CliInputError('<run_id> is required', 'Usage: moda run <run_id>');
-            }
-            const args = flagsToArgs(flags, 'run_id', positional);
-            const params = RunDetailSchema.parse(args);
-            const data = await callDataAPI(`/runs/${params.run_id}`);
-            context.output.writeData(data);
-            break;
-        }
         case 'loops': {
             const args = flagsToArgs(flags);
             const params = LoopsSchema.parse(args);
@@ -927,12 +866,6 @@ export const commandRegistry = createCommandRegistry([
         examples: ['moda failures', 'moda failures --tool lookupCustomer'],
         ...dataApiDefaults,
     }),
-    legacyCommand({
-        name: 'why-failed',
-        description: 'Explain a failed or partial production run',
-        examples: ['moda why-failed 12345678-1234-1234-1234-1234567890ab'],
-        ...dataApiDefaults,
-    }),
     legacyCommand({
         name: 'ask',
         description: 'Ask Moda a production intelligence question',
@@ -1014,24 +947,6 @@ export const commandRegistry = createCommandRegistry([
         examples: ['moda tool-failure-detail search --include-window'],
         ...dataApiDefaults,
     }),
-    legacyCommand({
-        name: 'workflows',
-        description: 'Recurring workflow catalog',
-        examples: ['moda workflows --time-range=7d'],
-        ...dataApiDefaults,
-    }),
-    legacyCommand({
-        name: 'runs',
-        description: 'List workflow executions',
-        examples: ['moda runs --time-range=24h --outcome=partial'],
-        ...dataApiDefaults,
-    }),
-    legacyCommand({
-        name: 'run',
-        description: 'Get a single workflow execution',
-        examples: ['moda run 12345678-1234-1234-1234-1234567890ab'],
-        ...dataApiDefaults,
-    }),
     legacyCommand({
         name: 'loops',
         description: 'List currently-active loop detections',

package/dist/production-intelligence.d.ts

@@ -5,10 +5,10 @@ declare const INVESTIGATION_SCHEMA_VERSION = "production_investigation.v0.1";
 declare const OVERVIEW_SCHEMA_VERSION = "production_overview.v0.1";
 declare const ASK_SCHEMA_VERSION = "production_ask.v0.1";
 type Severity = 'critical' | 'high' | 'medium' | 'low' | 'info';
-type FindingKind = 'tool_failure' | 'loop' | 'frustration' | 'partial_run' | 'prompt_regression' | 'skill_gap' | 'insufficient_data';
+type FindingKind = 'tool_failure' | 'loop' | 'frustration' | 'prompt_regression' | 'skill_gap' | 'insufficient_data';
 export interface ProductionEvidenceRef {
     id: string;
-    kind: 'data_api' | 'tool_failure' | 'frustration' | 'loop' | 'run' | 'harness' | 'ask';
+    kind: 'data_api' | 'tool_failure' | 'frustration' | 'loop' | 'harness' | 'ask';
     label: string;
     endpoint?: string;
     conversationId?: string;
@@ -52,7 +52,6 @@ export interface ProductionInvestigation {
     generatedAt: string;
     daysBack: number;
     scope: {
-        runId?: string;
         conversationId?: string;
         toolName?: string;
         loopId?: string;
@@ -136,7 +135,6 @@ export declare function loadOverviewBriefing(options: {
 export declare function investigateProduction(options: {
     daysBack: number;
     cwd: string;
-    runId?: string;
     conversationId?: string;
     toolName?: string;
     loopId?: string;

package/dist/production-intelligence.js

@@ -50,15 +50,6 @@ export async function runProductionCommand(context) {
         // it from the exit code alone (in addition to `data.degraded` / status).
         return { exitCode: result.degraded ? EXIT_DEGRADED : EXIT_OK };
     }
-    if (context.command === 'why-failed') {
-        const runId = context.positional;
-        if (!runId) {
-            throw new CliInputError('<run_id> is required', 'Usage: moda why-failed <run_id>');
-        }
-        const investigation = await investigateProduction({ daysBack, cwd: context.cwd, runId });
-        writeInvestigation(context, investigation);
-        return { exitCode: 0 };
-    }
     if (context.command === 'failures') {
         const investigation = await investigateProduction({
             daysBack,
@@ -73,7 +64,6 @@ export async function runProductionCommand(context) {
         const investigation = await investigateProduction({
             daysBack,
             cwd: context.cwd,
-            runId: context.flags.run,
             conversationId: context.flags.conversation,
             toolName: context.flags.tool,
             loopId: context.flags.loop,
@@ -137,20 +127,16 @@ export async function investigateProduction(options) {
         ? `/tool-failures/${encodeURIComponent(options.toolName)}?days_back=${options.daysBack}&limit=5`
         : `/tool-failures?days_back=${options.daysBack}`;
     const frustrationsEndpoint = `/frustrations?days_back=${options.daysBack}&limit=5`;
-    const runsEndpoint = options.runId
-        ? `/runs/${encodeURIComponent(options.runId)}`
-        : `/runs?time_range=${timeRangeForDays(options.daysBack)}&limit=20`;
     const loopsEndpoint = options.loopId
         ? `/loops/${encodeURIComponent(options.loopId)}`
         : '/loops/active?limit=20';
-    const [overview, tools, frustrations, runs, loops] = await Promise.all([
+    const [overview, tools, frustrations, loops] = await Promise.all([
         safeData(overviewEndpoint),
         safeData(toolEndpoint),
         safeData(frustrationsEndpoint),
-        safeData(runsEndpoint),
         safeData(loopsEndpoint),
     ]);
-    for (const [key, result] of Object.entries({ overview, tools, frustrations, runs, loops })) {
+    for (const [key, result] of Object.entries({ overview, tools, frustrations, loops })) {
         sourceStatus[key] = result.status;
         if (result.warning)
             warnings.push(result.warning);
@@ -161,11 +147,9 @@ export async function investigateProduction(options) {
     findings.push(...toolFailureFindings(tools.data, options.daysBack, hints, evidenceRefs, options.toolName));
     if (!options.failureOnly) {
         findings.push(...frustrationFindings(frustrations.data, options.daysBack, evidenceRefs));
-        findings.push(...runFindings(runs.data, evidenceRefs, options.runId));
         findings.push(...loopFindings(loops.data, evidenceRefs, options.loopId));
     }
     else {
-        findings.push(...runFindings(runs.data, evidenceRefs, options.runId).filter((finding) => finding.kind === 'partial_run'));
         findings.push(...loopFindings(loops.data, evidenceRefs, options.loopId));
     }
     if (findings.length === 0) {
@@ -179,7 +163,6 @@ export async function investigateProduction(options) {
         generatedAt: new Date().toISOString(),
         daysBack: options.daysBack,
         scope: {
-            ...(options.runId ? { runId: options.runId } : {}),
             ...(options.conversationId ? { conversationId: options.conversationId } : {}),
             ...(options.toolName ? { toolName: options.toolName } : {}),
             ...(options.loopId ? { loopId: options.loopId } : {}),
@@ -461,45 +444,6 @@ function frustrationFindings(data, daysBack, evidenceRefs) {
                 }],
         }];
 }
-function runFindings(data, evidenceRefs, scopedRunId) {
-    const rows = runRows(data);
-    return rows.flatMap((run) => {
-        const runId = run.id || scopedRunId;
-        const outcome = run.outcome;
-        if (!runId || !['failed', 'partial', 'abandoned', 'loop'].includes(outcome))
-            return [];
-        const evidenceId = `run:${runId}`;
-        evidenceRefs.push({
-            id: evidenceId,
-            kind: 'run',
-            label: `${run.agentName || 'Agent run'} ${outcome}`,
-            endpoint: `/runs/${runId}`,
-            runId,
-            conversationId: run.conversationId,
-            value: run.raw,
-        });
-        return [{
-                id: `finding_run_${slug(runId)}`,
-                kind: outcome === 'partial' ? 'partial_run' : 'tool_failure',
-                title: `${run.agentName || 'Agent run'} ended as ${outcome}`,
-                severity: outcome === 'failed' ? 'high' : 'medium',
-                confidence: 'high',
-                impactScore: outcome === 'failed' ? 88 : 76,
-                rankReason: `Run outcome is ${outcome}${run.reason ? `: ${run.reason}` : ''}.`,
-                summary: run.reason || 'Moda classified this run as incomplete or unhealthy.',
-                evidenceRefIds: [evidenceId],
-                likelyLocations: [unknownLocation('entrypoint', 'Map this run to a harness entrypoint after harness sync.')],
-                recommendedActions: [{
-                        id: `action_run_${slug(runId)}`,
-                        title: 'Open the run detail.',
-                        rationale: 'Run detail includes receipts and missing effects.',
-                        command: `moda run ${runId}`,
-                        mutability: 'read',
-                        requiresApproval: false,
-                    }],
-            }];
-    });
-}
 function loopFindings(data, evidenceRefs, scopedLoopId) {
     const rows = loopRows(data);
     return rows.flatMap((loop) => {
@@ -1116,24 +1060,6 @@ function toolRows(data, scopedTool) {
             }];
     });
 }
-function runRows(data) {
-    const record = asRecord(data);
-    const run = asRecord(record?.run);
-    const rows = Array.isArray(record?.runs) ? record.runs : run ? [run] : [];
-    return rows.flatMap((item) => {
-        const row = asRecord(item);
-        if (!row)
-            return [];
-        return [{
-                id: asString(row.job_id) ?? asString(row.run_id) ?? '',
-                outcome: asString(row.outcome) ?? 'unknown',
-                agentName: asString(row.agent_name) ?? undefined,
-                reason: asString(row.outcome_reason) ?? undefined,
-                conversationId: asString(row.conversation_id) ?? undefined,
-                raw: row,
-            }];
-    });
-}
 function loopRows(data) {
     const record = asRecord(data);
     const rows = Array.isArray(record?.loops)
@@ -1397,17 +1323,6 @@ function numberFlag(flags, key, fallback) {
         return fallback;
     return Math.min(Math.floor(parsed), 90);
 }
-function timeRangeForDays(days) {
-    if (days <= 1)
-        return '24h';
-    if (days <= 3)
-        return '3d';
-    if (days <= 7)
-        return '7d';
-    if (days <= 30)
-        return '30d';
-    return '90d';
-}
 function normalizeConfidence(value) {
     if (value === 'high' || value === 'medium' || value === 'low' || value === 'unknown')
         return value;
@@ -1421,7 +1336,6 @@ function evidenceKind(value) {
         value === 'tool_failure' ||
         value === 'frustration' ||
         value === 'loop' ||
-        value === 'run' ||
         value === 'harness' ||
         value === 'ask')
         return value;
@@ -1443,8 +1357,6 @@ function purposeForCommand(command) {
         return 'Read the relevant conversation window.';
     if (command.startsWith('moda loop'))
         return 'Inspect loop detection details.';
-    if (command.startsWith('moda run'))
-        return 'Inspect one agent run and side-effect receipts.';
     if (command.startsWith('moda doctor'))
         return 'Validate Moda setup and data flow.';
     if (command.startsWith('moda overview'))

package/dist/schemas.d.ts

@@ -185,49 +185,6 @@ export declare const SearchSchema: z.ZodObject<{
     limit?: number | undefined;
     user_id?: string | undefined;
 }>;
-/** moda workflows — recurring workflow catalog (read from agent_job_classes). */
-export declare const WorkflowsSchema: z.ZodObject<{
-    time_range: z.ZodOptional<z.ZodDefault<z.ZodEnum<["all", "1h", "3d", "7d", "24h", "30d", "90d"]>>>;
-    user_id: z.ZodOptional<z.ZodString>;
-}, "strip", z.ZodTypeAny, {
-    time_range?: "all" | "1h" | "3d" | "7d" | "24h" | "30d" | "90d" | undefined;
-    user_id?: string | undefined;
-}, {
-    time_range?: "all" | "1h" | "3d" | "7d" | "24h" | "30d" | "90d" | undefined;
-    user_id?: string | undefined;
-}>;
-/** moda runs — list workflow executions with classified outcomes. */
-export declare const RunsSchema: z.ZodObject<{
-    time_range: z.ZodOptional<z.ZodDefault<z.ZodEnum<["all", "1h", "3d", "7d", "24h", "30d", "90d"]>>>;
-    outcome: z.ZodOptional<z.ZodEnum<["running", "succeeded", "partial", "failed", "loop", "abandoned", "unknown"]>>;
-    /** Filter to one workflow class id (e.g. cron:helpscout-triage-hourly). */
-    class_id: z.ZodOptional<z.ZodString>;
-    user_id: z.ZodOptional<z.ZodString>;
-    limit: z.ZodOptional<z.ZodDefault<z.ZodNumber>>;
-    offset: z.ZodOptional<z.ZodDefault<z.ZodNumber>>;
-}, "strip", z.ZodTypeAny, {
-    time_range?: "all" | "1h" | "3d" | "7d" | "24h" | "30d" | "90d" | undefined;
-    limit?: number | undefined;
-    offset?: number | undefined;
-    user_id?: string | undefined;
-    outcome?: "unknown" | "loop" | "running" | "failed" | "succeeded" | "partial" | "abandoned" | undefined;
-    class_id?: string | undefined;
-}, {
-    time_range?: "all" | "1h" | "3d" | "7d" | "24h" | "30d" | "90d" | undefined;
-    limit?: number | undefined;
-    offset?: number | undefined;
-    user_id?: string | undefined;
-    outcome?: "unknown" | "loop" | "running" | "failed" | "succeeded" | "partial" | "abandoned" | undefined;
-    class_id?: string | undefined;
-}>;
-/** moda run <runId> — single execution detail with side-effect receipts. */
-export declare const RunDetailSchema: z.ZodObject<{
-    run_id: z.ZodString;
-}, "strip", z.ZodTypeAny, {
-    run_id: string;
-}, {
-    run_id: string;
-}>;
 /** moda loops — currently active loop detections. */
 export declare const LoopsSchema: z.ZodObject<{
     limit: z.ZodOptional<z.ZodDefault<z.ZodNumber>>;

package/dist/schemas.js

@@ -78,13 +78,6 @@ export const AuditSchema = z.object({
     include_raw: boolFlag(),
     limit: z.number().min(1).max(1000).default(200).optional(),
 });
-// ─── Workflows + Runs (agent runs runtime) ──────────────────
-//
-// Naming model:
-//   workflow = the recurring class definition (e.g. cron:helpscout-triage-hourly)
-//   run      = one execution of a workflow
-// `moda workflows` lists the catalog; `moda runs` lists executions.
-const RUN_OUTCOMES = ['running', 'succeeded', 'partial', 'failed', 'loop', 'abandoned', 'unknown'];
 const TIME_RANGES = ['all', '1h', '3d', '7d', '24h', '30d', '90d'];
 const SEARCH_MODES = ['keyword', 'semantic', 'hybrid'];
 /**
@@ -99,25 +92,6 @@ export const SearchSchema = z.object({
     time_range: z.enum(TIME_RANGES).default('all').optional(),
     limit: z.number().min(1).max(100).default(20).optional(),
 });
-/** moda workflows — recurring workflow catalog (read from agent_job_classes). */
-export const WorkflowsSchema = z.object({
-    time_range: z.enum(TIME_RANGES).default('all').optional(),
-    user_id: z.string().optional(),
-});
-/** moda runs — list workflow executions with classified outcomes. */
-export const RunsSchema = z.object({
-    time_range: z.enum(TIME_RANGES).default('7d').optional(),
-    outcome: z.enum(RUN_OUTCOMES).optional(),
-    /** Filter to one workflow class id (e.g. cron:helpscout-triage-hourly). */
-    class_id: z.string().optional(),
-    user_id: z.string().optional(),
-    limit: z.number().min(1).max(100).default(50).optional(),
-    offset: z.number().min(0).default(0).optional(),
-});
-/** moda run <runId> — single execution detail with side-effect receipts. */
-export const RunDetailSchema = z.object({
-    run_id: z.string().uuid({ message: 'run_id must be a UUID' }),
-});
 /** moda loops — currently active loop detections. */
 export const LoopsSchema = z.object({
     limit: z.number().min(1).max(200).default(25).optional(),

package/dist/skills.js

@@ -27,6 +27,7 @@ import { CliInputError } from './errors.js';
 const SKILLS_REQUEST_TIMEOUT_MS = 15_000;
 const VALID_STATUSES = new Set(['approved', 'proposed', 'all']);
 const SKILLS_MANIFEST_PATH = join('.moda', 'skills.yml');
+const SKILL_PROPOSALS_MANIFEST_PATH = join('.moda', 'skill-proposals.yml');
 /**
  * Fetch a tenant's generated skills from the ingest worker. Uses the persisted
  * API key + ingest URL (`~/.moda/config.json`), mirroring snapshot auth.
@@ -342,11 +343,96 @@ function getGitCommit(cwd) {
         return undefined;
     }
 }
+async function fetchSkillProposals(flags, profileOptions = {}) {
+    const params = new URLSearchParams();
+    if (flags['generation-run-id'])
+        params.set('generationRunId', flags['generation-run-id']);
+    const status = flags.status ?? 'ready_for_pr';
+    if (status !== 'all')
+        params.set('status', status);
+    if (flags['tenant-id'])
+        params.set('tenantId', flags['tenant-id']);
+    const suffix = params.toString() ? `?${params.toString()}` : '';
+    const response = (await callControlAPI(`/skills/proposals${suffix}`, { method: 'GET' }, profileOptions));
+    return Array.isArray(response.proposals) ? response.proposals : [];
+}
+async function applySkillProposal(proposalId, flags, profileOptions = {}, options = {}) {
+    const cwd = options.cwd ?? process.cwd();
+    const params = new URLSearchParams();
+    if (flags['tenant-id'])
+        params.set('tenantId', flags['tenant-id']);
+    const suffix = params.toString() ? `?${params.toString()}` : '';
+    const proposal = (await callControlAPI(`/skills/proposals/${encodeURIComponent(proposalId)}${suffix}`, { method: 'GET' }, profileOptions));
+    if (!proposal.skillMd?.trim()) {
+        throw new CliInputError(`Proposal ${proposalId} does not include SKILL.md content`);
+    }
+    const safeId = safeSkillId(proposal.skillId);
+    const relativePath = `.claude/skills/${safeId}/SKILL.md`;
+    const writeStatus = writeIfChanged(join(cwd, relativePath), proposal.skillMd);
+    if (writeStatus === 'failed') {
+        throw new Error(`Failed to write ${relativePath}`);
+    }
+    writeSkillProposalManifest(cwd, proposal, relativePath);
+    let acknowledged = true;
+    let ackError;
+    try {
+        await callControlAPI(`/skills/proposals/${encodeURIComponent(proposalId)}/mark-applied${suffix}`, {
+            method: 'POST',
+            body: JSON.stringify({ appliedBy: 'moda-cli' }),
+        }, profileOptions);
+    }
+    catch (err) {
+        acknowledged = false;
+        ackError = err instanceof Error ? err.message : String(err);
+    }
+    return {
+        proposalId,
+        skillId: proposal.skillId,
+        path: relativePath,
+        status: writeStatus,
+        acknowledged,
+        ackError,
+    };
+}
+function safeSkillId(skillId) {
+    const safeId = basename(skillId || '');
+    if (!safeId || !/^[A-Za-z0-9_-]+$/.test(safeId)) {
+        throw new CliInputError(`Invalid skill_id in proposal: ${JSON.stringify(skillId).slice(0, 80)}`);
+    }
+    return safeId;
+}
+function writeSkillProposalManifest(cwd, proposal, relativePath) {
+    const manifestPath = join(cwd, SKILL_PROPOSALS_MANIFEST_PATH);
+    mkdirSync(dirname(manifestPath), { recursive: true });
+    const existing = existsSync(manifestPath) ? readFileSync(manifestPath, 'utf8') : 'applied_proposals:\n';
+    const entry = [
+        `  - proposal_id: "${escapeYamlString(proposal.proposalId)}"`,
+        `    skill_id: "${escapeYamlString(proposal.skillId)}"`,
+        `    skill_family_id: "${escapeYamlString(proposal.skillFamilyId)}"`,
+        `    skill_version: ${Number(proposal.skillVersion || 0)}`,
+        `    generation_run_id: "${escapeYamlString(proposal.generationRunId)}"`,
+        `    clustering_run_id: "${escapeYamlString(proposal.clusteringRunId)}"`,
+        `    topic_id: "${escapeYamlString(proposal.topicId)}"`,
+        `    replay_set_id: "${escapeYamlString(proposal.replaySetId)}"`,
+        `    delta_pass_rate: ${Number(proposal.deltaPassRate || 0)}`,
+        `    path: "${escapeYamlString(relativePath)}"`,
+        `    applied_at: "${new Date().toISOString()}"`,
+    ].join('\n');
+    const prefix = existing.endsWith('\n') ? existing : `${existing}\n`;
+    writeFileSync(manifestPath, `${prefix}${entry}\n`);
+}
+function escapeYamlString(value) {
+    return String(value ?? '').replace(/\\/g, '\\\\').replace(/"/g, '\\"');
+}
+function formatProposalRate(value) {
+    return `${(Number(value || 0) * 100).toFixed(1)}%`;
+}
 /**
  * `moda skills <subcommand>` dispatcher.
  */
 export async function runSkillsCommand(args) {
     const sub = args.positional;
+    const positionals = args.positionals ?? (sub ? [sub] : []);
     if (sub === 'gen') {
         const result = await generateSkills(args.flags);
         if (args.flags.json === 'true') {
@@ -371,13 +457,63 @@ export async function runSkillsCommand(args) {
         }
         return Number(result.sessions_failed ?? 0) > 0 ? 1 : 0;
     }
+    if (sub === 'proposals' && positionals[1] === 'list') {
+        const proposals = await fetchSkillProposals(args.flags, { profile: args.flags.profile });
+        if (args.flags.json === 'true') {
+            console.log(JSON.stringify({ proposals }));
+        }
+        else if (proposals.length === 0) {
+            console.log('No skill proposals found.');
+        }
+        else {
+            for (const proposal of proposals) {
+                console.log([
+                    proposal.proposalId,
+                    proposal.proposalStatus,
+                    `${proposal.skillId}@v${proposal.skillVersion}`,
+                    `delta=${formatProposalRate(proposal.deltaPassRate)}`,
+                    `baseline=${formatProposalRate(proposal.baselinePassRate)}`,
+                    `candidate=${formatProposalRate(proposal.candidatePassRate)}`,
+                    proposal.name,
+                ].filter(Boolean).join('\t'));
+            }
+        }
+        return 0;
+    }
+    if (sub === 'proposal' && positionals[1] === 'apply') {
+        const proposalId = positionals[2] || args.flags['proposal-id'];
+        if (!proposalId) {
+            throw new CliInputError('A proposal id is required', 'Usage: moda skills proposal apply <proposal-id>');
+        }
+        const result = await applySkillProposal(proposalId, args.flags, { profile: args.flags.profile });
+        if (args.flags.json === 'true') {
+            console.log(JSON.stringify(result));
+        }
+        else {
+            console.log([
+                `${result.status} ${result.path}`,
+                result.acknowledged ? 'acknowledged: true' : 'acknowledged: false',
+            ].join('\n'));
+        }
+        if (!result.acknowledged) {
+            // The local SKILL.md was written, but the server still lists this
+            // proposal as ready_for_pr. Warn (to stderr so JSON stdout stays clean)
+            // and exit non-zero so the failed acknowledgement isn't silently treated
+            // as success — re-running `apply` re-attempts the acknowledgement.
+            console.error(`Warning: wrote ${result.path} but failed to acknowledge proposal ${proposalId} on the server` +
+                (result.ackError ? ` (${result.ackError})` : '') +
+                '. The proposal is still listed as ready — re-run `moda skills proposal apply` to acknowledge it.');
+            return 1;
+        }
+        return 0;
+    }
     if (sub === 'sync') {
         const result = await runSkillSync(args.flags, { profile: args.flags.profile });
         console.log(JSON.stringify({ synced: result.synced, dryRun: result.dryRun }));
         return 0;
     }
     if (sub !== 'pull') {
-        throw new CliInputError(sub ? `Unknown skills subcommand '${sub}'` : 'A skills subcommand is required', 'Usage: moda skills gen [--source=all|sdk|coding-agent] [--max-sessions=N] [--start-at=ISO] [--end-at=ISO] [--replay] [--reprocess-segments] [--force-reprocess] [--wait] | moda skills pull [--status=approved|proposed|all] | moda skills sync [--dry-run]');
+        throw new CliInputError(sub ? `Unknown skills subcommand '${sub}'` : 'A skills subcommand is required', 'Usage: moda skills gen [--source=all|sdk|coding-agent] [--max-sessions=N] [--start-at=ISO] [--end-at=ISO] [--replay] [--replay-set-id=ID] [--reprocess-segments] [--force-reprocess] [--wait] | moda skills proposals list [--status=ready_for_pr] | moda skills proposal apply <proposal-id> | moda skills pull [--status=approved|proposed|all] | moda skills sync [--dry-run]');
     }
     const status = (args.flags.status ?? 'approved').toLowerCase();
     if (!VALID_STATUSES.has(status)) {
@@ -418,6 +554,7 @@ async function generateSkills(flags) {
         start_at: flags['start-at'] || undefined,
         end_at: flags['end-at'] || undefined,
         clustering_run_id: flags['clustering-run-id'] || undefined,
+        replay_set_id: flags['replay-set-id'] || flags['eval-set-id'] || undefined,
         run_clustering: flags['no-clustering'] !== 'true',
         source: normalizeSkillSource(flags.source),
         clustering_timeout_seconds: parsePositiveInt(flags['clustering-timeout-seconds'], 0),

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@moda-ai/cli",
-  "version": "1.10.0",
+  "version": "1.10.2",
   "description": "CLI for Moda - AI agent analytics and observability",
   "type": "module",
   "bin": {

package/skills/moda-cli/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: moda-cli
 version: 2.0.0
-description: Query Moda's AI agent conversation analytics and manage code-first prompt versions from the terminal — semantic/keyword/hybrid message search, overview KPIs, topic clusters, message context, user frustration detections, tool failures, agent jobs, recurring workflows, active loop detections, and moda prompts status/sync/promote. Use when the user asks about moda, modaflows, conversation analytics, prompt management, user frustrations, agent observability, tool failure debugging, why an agent run partially completed or looped, wants to find conversations or tool calls about a topic, or wants to investigate how their AI agent is performing.
+description: Query Moda's AI agent conversation analytics and manage code-first prompt versions from the terminal — semantic/keyword/hybrid message search, overview KPIs, topic clusters, message context, user frustration detections, tool failures, active loop detections, and moda prompts status/sync/promote. Use when the user asks about moda, modaflows, conversation analytics, prompt management, user frustrations, agent observability, tool failure debugging, wants to find conversations or tool calls about a topic, or wants to investigate how their AI agent is performing.
 ---
 
 # Moda CLI
@@ -369,42 +369,7 @@ Pass `--include-window` to attach a `window` field per row with the message
 slice centered on the anchor. `--window=N` (1–5, default `1`) is the
 half-width; `--window=1` ⇒ 3 messages total (anchor ± 1).
 
-### 7. Workflow runs and outcomes
-
-Every triggered run of an agent (cron invocation, scheduled task, or
-interactive session) is classified `succeeded | partial | failed | loop |
-abandoned`. Use this layer to answer "did the job actually do its work?",
-which is structurally different from "did it error?" — most jobs that
-partially completed never threw.
-
-```bash
-moda runs                                       # last 7d, all outcomes
-moda runs --time-range=24h --outcome=partial    # which runs only half-finished
-moda runs --outcome=loop                        # runs caught in active loops
-moda runs --class-id=cron:helpscout-triage-hourly --limit=20
-moda run <run_id>                               # one execution + its side-effect receipts
-```
-
-`moda run <run_id>` returns the `agent_jobs` row plus every
-`side_effect_receipts` entry for the run — i.e. every tool call the agent
-made (read / write / send / spawn / compute), with success status and
-target system. This is the agent-as-employee log; use it instead of (or
-alongside) `moda context` when debugging "what did the agent actually
-DO".
-
-### 8. Recurring workflows / cron status board
-
-```bash
-moda workflows                                  # health per recurring class
-moda workflows --time-range=7d                  # filter to last 7d of activity
-```
-
-Each row is one job class (a cron slug, idea-widget id, or sourceType)
-with 30d rolling success / partial / failed / loop counts and a
-`health_score` (0–1). Sort by health to find degraded customer-facing
-crons that aren't crashing but aren't completing either.
-
-### 9. Active loop investigation
+### 7. Active loop investigation
 
 ```bash
 moda loops                                      # everything currently stuck
@@ -418,7 +383,7 @@ agent re-ran the same action 378 times. Use the `loop_signature_human`
 field to recognize cross-conversation patterns (one tool definition bug
 will produce identical signatures across many users).
 
-### 10. Prompt management
+### 8. Prompt management
 
 Use these only when working on the code-first prompt registry.
 
@@ -505,19 +470,6 @@ moda clusters --parent-id=<node_id>
 moda cluster-conversations <node_id>
 ```
 
-### Find degraded crons before customers do
-
-```bash
-# Health-sorted workflow board
-moda workflows | jq '.workflows | sort_by(.health) | .[] | select(.runs >= 4 and .health < 0.85)'
-
-# What loops are eating tokens right now?
-moda loops --limit=20 | jq '.loops[] | {sig: .loop_signature_human, visits: .cycle_visits, conv: .conversation_id}'
-
-# Drill into a partial run
-moda runs --outcome=partial --limit=5 | jq -r '.runs[].job_id' | head -1 | xargs -I{} moda run {}
-```
-
 ### One-liner chains with jq
 
 ```bash
@@ -596,9 +548,6 @@ npx: `npx -p @moda-ai/cli moda <command>`.
 | `moda frustrations` | User frustration detections with evidence |
 | `moda tool-failures` | Tool failure overview |
 | `moda tool-failure-detail <tool_name>` | Per-tool failure breakdown + examples |
-| `moda workflows` | Recurring workflow catalog with 30d health per class |
-| `moda runs` | Recent workflow executions with classified outcomes |
-| `moda run <run_id>` | One execution + its side-effect receipts |
 | `moda loops` | Currently-active loop detections |
 | `moda loop <detection_id>` | One loop signature with blast radius |
 | `moda prompts status` | Read-only local prompt status |
@@ -610,8 +559,7 @@ npx: `npx -p @moda-ai/cli moda <command>`.
 `--user-id`, `--time-range=<range>`, `--limit=N` (1–100, default 20).
 
 Common flags across most commands: `--days-back=N` (1–90), `--limit=N`,
-`--offset=N`, `--time-range=<range>`. Runs commands also accept
-`--outcome=<outcome>`, `--class-id=<workflow_class_id>`, and `--user-id=<id>`.
+`--offset=N`, `--time-range=<range>`.
 `frustrations` and `tool-failure-detail` additionally accept
 `--include-window` (attach a `window` slice per row) and `--window=N` (1–5,
 default `1`; half-width, so `--window=1` ⇒ 3 messages).