Shreyas70773
UserSafety-first agentic Oracle DBA skills for Claude Code: 24 tier-gated executable domains (RMAN, Data Guard, RAC, perf, security…) fused with Oracle's official knowledge base, native SQLcl-MCP + a custom guarded host-ops MCP server.
Categories
Indexed Skills (25)
oracle-dba-super
USE FOR ANY Oracle Database request — whether to UNDERSTAND/DESIGN/WRITE (knowledge) or to OPERATE/CHANGE/RECOVER a database (action). Covers SQL & PL/SQL authoring, tuning, schema design, migrations (from PostgreSQL/MySQL/SQL Server/Mongo/Snowflake), JDBC & frameworks, ORDS, SQLcl/MCP, AND full DBA operations: provisioning, backup & recovery (RMAN/Data Pump), patching & upgrades, RAC, Data Guard, GoldenGate, performance/AWR/ASH, space/ASM, users/ roles/privileges, security/TDE/auditing, PDB/multitenant, monitoring/alerting, jobs, flashback, capacity, networking/listener, cloning/refresh, Exadata, OEM, OCI/Autonomous, decommissioning, and ANY ORA-/RMAN-/CRS- error. Trigger even if the user does not name a task or tool. This is the MASTER ROUTER: it sends knowledge questions to the vendored Oracle knowledge base and operational requests to the right tier-gated action domain, always loading the agent-safety patterns and enforcing the autonomy policy (T1 self-exec / T2 dry-run+execute / T3 runbook + human approv
oracle-dba-common
MANDATORY shared foundation for EVERY oracle-dba skill. Import this BEFORE writing or running any DBA task. It is the single source of truth for: the SQLcl MCP connection contract (saved named connection "dba_ai_conn" only — never SYS/SYSTEM, never hardcoded creds), secret resolution from the Oracle Wallet / external password store (NO plaintext in argv/env/files EVER), the structured logging contract and canonical log paths, the exit-code convention (0=ok, documented non-zero codes), and the risk-tier banner + autonomy policy (T1 observe / T2 reversible-with-dry-run / T3 destructive-runbook-only). Do NOT reinvent connection, logging, dry-run, or approval-token logic in any other skill — source _common/scripts/lib.sh or lib.ps1 and follow this contract verbatim.
oracle-dba-advanced-features
USE THIS SKILL WHENEVER the user mentions Oracle Advanced Compression / OLTP / Advanced Row compression / ROW STORE COMPRESS ADVANCED, Hybrid Columnar Compression (HCC) / COLUMN STORE COMPRESS FOR QUERY|ARCHIVE / Exadata-qualified compression, Database In-Memory / INMEMORY / IM column store / INMEMORY_SIZE / V$IM_SEGMENTS / MEMCOMPRESS, Partitioning implementation / RANGE|LIST|HASH|INTERVAL|composite / partition add-drop- split-merge / heap-to-partitioned conversion / partitioning strategy, Oracle Sharding / GDSCTL / GSM / shard director / shardcatalog / sharded & duplicated tables / chunks, or Result Cache / RESULT_CACHE_MODE / RESULT_CACHE_MAX_SIZE / DBMS_RESULT_CACHE — EVEN IF they do not name the specific task. Covers (T1) compression/HCC/In-Memory/partition/ Result-Cache/sharding INVENTORY & advisor reports; (T2, --dry-run default) OLTP-compress enable/disable, HCC attribute set, In-Memory per-object enable/disable, partition add/ drop on an already-partitioned table, and Result Cache configure; and (T3,
oracle-dba-backup-and-recovery
USE THIS SKILL WHENEVER the user mentions RMAN, backups, restore, recover/recovery, Data Pump, expdp/impdp, archivelog, FRA / fast recovery area, retention policy, crosscheck, obsolete/expired backups, recovery catalog, control file or spfile autobackup, restore points, flashback, PITR / point-in-time recovery, TSPITR, block media recovery / block corruption, DR drills, cross-platform CONVERT, or any of the errors ORA-01578 (block corruption), ORA-19xxx (RMAN/backup I/O), or RMAN-xxxxx — EVEN IF they do not name the specific task. Covers scheduled RMAN full/incremental/archivelog backups, RMAN CONFIGURE, backup status monitoring and VALIDATE, crosscheck+delete within retention, recovery-catalog maintenance, Data Pump export/import, guaranteed restore point create/list/drop, and (plan-only, human-driven) recovery runbooks. Follows the oracle-dba-common contracts: SQLcl MCP saved connection dba_ai_conn, Oracle Wallet secrets (never plaintext), --dry-run defaults on T2, and tier gating (max self-execute = T2; al
oracle-dba-capacity-planning-reporting
USE THIS SKILL WHENEVER the user asks about capacity planning, capacity/sizing reviews, growth trend analysis or forecasting, "days to full" / "when will this tablespace fill", space-usage trend reports, database/instance status reports (daily/weekly health snapshot), resource-utilization reports (CPU / memory / IO / sessions / AAS), SLA / uptime / availability reporting, chargeback or showback / cost-allocation reporting, or a capacity-planning recommendation — EVEN IF they do not name the specific report. Covers the read-only reporting family (status, space-trend, resource, SLA, growth forecast, capacity recommendation, chargeback) plus the single additive daily space-metrics capture that feeds long-term forecasting. Follows the oracle-dba-common contracts: SQLcl MCP saved connection dba_ai_conn, Oracle Wallet secrets (never plaintext), --dry-run default on the one T2 mechanism, and tier gating (max self-execute = T2; all PROVISIONING and repo/structural DDL is T3 runbook).
oracle-dba-cloning-refresh-migration
USE THIS SKILL for ANY Oracle cloning, environment refresh, or migration request — even when the user does NOT name the task or tool. Covers: RMAN DUPLICATE / database clone, PDB clone & refreshable clones (hot/remote-over-DB-link), snapshot / storage (ASM/NFS/array) clones via hot-backup mode, Data Pump schema refresh (impdp REMAP), test/dev refresh orchestration + non-prod data masking/subsetting, classic transportable tablespaces (TTS), cross-platform transportable tablespaces (XTTS) + endian conversion, Zero Downtime Migration (ZDM) to cloud, and heterogeneous (non-Oracle Gateway / HS / GoldenGate-staging) migration readiness. TRIGGER on symptoms and codes too: "clone prod to test", "refresh QA from prod", "duplicate the database", "stand up a copy", "move to OCI/cloud", "go-live cutover", "transport these tablespaces", "different endian platform", "mask PII in the copy"; and on ORA-19505 / ORA-19625 (DUPLICATE file issues), ORA-65261/65000 (refreshable PDB), ORA-39001/39002/ 39082/39083/31626/31693 (Data
oracle-dba-cloud-autonomous-oci
USE THIS SKILL WHENEVER the user mentions Oracle Cloud / OCI, Autonomous Database (ADB / ATP / ADW / AJD / APEX), Autonomous DB monitoring or auto-patching, OCI Base Database Service / DB systems, cloud-managed backups / backup retention / long-term backups, ECPU/OCPU or storage auto-scaling, OCI IAM / database-access policy, the OCI CLI (`oci db ...`, `oci iam ...`), Exadata Cloud@Customer (ExaCC) / ExaDB-C / VM clusters, or cloud migration via Zero Downtime Migration (ZDM / zdmcli) or the OCI Database Migration Service (DMS) — EVEN IF they do not name the specific task. Covers Autonomous DB & Base DB provisioning (T2, dry-run-default), cloud backup configuration, auto-scaling configuration, OCI IAM policy-as-code (create/update), read-only ADB health and auto-patch monitoring, and the (plan-only, human-driven) Tier-3 runbooks for ExaCC management and cloud migration. Follows the oracle-dba-common contracts: SQLcl MCP saved connection dba_ai_conn for in-DB work, OCI CLI for control-plane host ops, Oracle Wal
oracle-dba-data-guard-disaster-recovery
USE THIS SKILL for ANY Oracle Data Guard / standby / disaster-recovery / HA request - EVEN IF the user does not name the task. Triggers: physical/standby database, redo transport & apply, apply/transport LAG, archive/redo GAP, switchover, FAILOVER, reinstate failed primary, standby build/rebuild (RMAN active duplicate), Active Data Guard / real-time query / DML redirect, Fast-Start Failover (FSFO) + observer, snapshot standby, cascaded standby, Data Guard broker (dgmgrl), protection modes (Max Protection/Availability/ Performance), Far Sync, role-based services. Symptoms: "standby is behind", "apply stopped", "MRP not running", "redo not shipping", "lost the primary", "DR site", "promote standby". Error codes: ORA-16xxx broker/DG family (ORA-16700/16766/16778/16789/16809/16810/16664/16661 reinstate, ORA-16143/16131), ORA-01153, ORA-19909 (standby_became_primary), ORA-00752, RMAN-05xxx during DUPLICATE ... FOR STANDBY, CRS-xxxx when srvctl relocates DG services. Follows oracle-dba-common contracts: in-DB work
oracle-dba-decommissioning-lifecycle-end
USE THIS SKILL WHENEVER the user is retiring, decommissioning, sunsetting, parking, or ending the life of an Oracle database, PDB, schema, or instance - EVEN IF they do not name the specific task. Covers the FINAL backup before decommission (RMAN KEEP archival), orderly database shutdown / park (SHUTDOWN IMMEDIATE / srvctl stop), schema / object cleanup (reversible ACCOUNT LOCK + PASSWORD EXPIRE quarantine, and the irreversible DROP USER runbook), license / inventory reclamation evidence, data archival & retention compliance (logical archive export + ILM / Flashback Data Archive posture + governed purge), and secure data destruction (TDE crypto-erase / OS secure wipe / media shred). Trigger phrases: "decommission", "retire / sunset the database", "shut down for good", "park the DB", "drop the database / schema", "data destruction / sanitize / wipe", "retention / legal hold at end of life", "reclaim the license", "end of life", "tear down". Follows the oracle-dba-common contracts: SQLcl MCP saved connection db
oracle-dba-enterprise-manager-oem-tooling
USE THIS SKILL WHENEVER the user mentions Oracle Enterprise Manager (OEM), Cloud Control, EM 13.5 / EM 24ai, the OMS (Oracle Management Service), the Management Repository / SYSMAN schema, emcli or emctl, a management agent, OEM targets, blackouts, monitoring templates, metric/threshold templates, OEM repository administration, or an OEM / Cloud Control upgrade — EVEN IF they do not name the specific task. Covers OEM target/agent management (add/sync/relocate/remove), blackout management (create/stop), monitoring template management (export/create/apply/delete), repository administration (status, SYSMAN stats gather, pause/resume repository collection jobs), and the plan-only Tier-3 OEM/Cloud Control upgrade + repository DR + destructive-repo runbooks. Follows the oracle-dba-common contracts: SQLcl MCP saved connection dba_ai_conn, Oracle Wallet secrets (never plaintext), --dry-run defaults on T2, and tier gating (max self-execute = T2; the OEM upgrade and destructive repository ops are T3).
oracle-dba-exadata-engineered-systems
USE THIS SKILL WHENEVER the user mentions Exadata, Exadata Database Machine, Engineered Systems, storage cells / cellcli / dcli / griddisk / celldisk / physicaldisk, Smart Scan / cell offload / storage index, Exadata Smart Flash Cache / flash log / CELL_FLASH_CACHE, IORM / IO Resource Manager / IORMPLAN, exachk / AHF / TFA on Exadata, patchmgr / Exadata patching (cells, RoCE/InfiniBand switches, compute/db nodes, image/RU), Exadata expansion / rack scale-out / add cells / ASM rebalance on Exadata, Exadata Cloud@Customer storage cells, or Exadata-side ASM disk-group / failgroup work — EVEN IF they do not name the specific task. Covers (T1) exachk/AHF health + in-DB cell visibility, cell-status LIST, IORM/DBRM read, Smart Scan / flash / expansion diagnostics; (T2, --dry-run default) patchmgr PRECHECK, in-DB DBRM plan activation, and the CELL_FLASH_CACHE segment hint; and (T3, runbook-only) cell/switch/compute patch APPLY, storage-cell management, cell IORMPLAN configuration, cell flash-cache tuning, and rack ex
oracle-dba-flashback-technologies
USE THIS SKILL WHENEVER the user mentions Oracle Flashback in any form — Flashback Query / AS OF, Flashback Version Query / VERSIONS BETWEEN, Flashback Table, Flashback Drop / recyclebin / "undrop" a dropped table, Flashback Transaction / TRANSACTION_BACKOUT / backing out a committed transaction, restore points (normal or GUARANTEE FLASHBACK DATABASE), Flashback Data Archive / FDA / "Total Recall" / long-term row history, or Flashback Database / FLASHBACK DATABASE TO RESTORE POINT / rewinding the whole database — EVEN IF they do not name the exact task. Also triggers on undo-based "get the old rows back", "recover a dropped/truncated table without a restore", or errors ORA-01555 (snapshot too old), ORA-08180 (no snapshot at this time), ORA-38xxx (flashback/restore point), ORA-55xxx (flashback archive / transaction backout). Covers read-only inspection (Flashback Query/Version Query, recyclebin/FDA/restore-point listing), reversible T2 recovery (Flashback Table, Flashback Drop undrop, Flashback Transaction bac
oracle-dba-goldengate-replication
USE THIS for ANY Oracle GoldenGate / logical-replication / materialized-view replication request - EVEN IF the user does not name a task or tool. Covers Extract / Replicat / Distribution-path / Manager status, START/STOP/RESTART, replication LAG (Integrated apply latency, capture latency, heartbeat lag), the automatic heartbeat table (ADD HEARTBEATTABLE / GG_LAG / GGS_HEARTBEAT), trail-file inventory + purge/retention, initial-load (instantiation SCN, HANDLECOLLISIONS, AFTERCSN, expdp/impdp parfiles), supplemental-logging readiness, Conflict Detection & Resolution (CDR / COMPARECOLS / RESOLVECONFLICT), logical-replication troubleshooting (missing rows, lag, ABENDED process), and materialized-view replication refresh/staleness. Trigger on symptoms too: "replication is behind / lagging", "rows not arriving on target", "Extract abended", "Replicat stopped", "OGG-xxxxx", "ggsci / adminclient", "trail files filling the disk", "heartbeat stale", "MV won't refresh", "ORA-26xxx" (LogMiner / Streams / GoldenGate apply
oracle-dba-job-scheduling-automation
USE THIS SKILL WHENEVER the user mentions DBMS_SCHEDULER, the Oracle Scheduler, scheduler jobs / programs / schedules / job classes / windows, job CHAINS, external (OS-executable) jobs, scheduler CREDENTIALS, calendaring repeat_interval strings, cron jobs or Windows Task Scheduler that drive Oracle work, BROKEN / FAILED / stuck scheduler jobs, rescheduling a job, "run a job now" / STOP_JOB / RUN_JOB, job-log retention/purge, or migrating a legacy DBMS_JOB — EVEN IF they do not name the specific task. Covers scheduler + OS-cron job MONITORING (T1), DBMS_SCHEDULER and external job CREATION, RESCHEDULING, job-chain configuration, and job-failure diagnose+remediate (T2, dry-run default), plus the plan-only Tier-3 runbooks for drops / forced runs / complex chains / credential creation / log purge / DBMS_JOB migration. Follows the oracle-dba-common contracts: SQLcl MCP saved connection dba_ai_conn, Oracle Wallet secrets (never plaintext, never SYS/SYSTEM by default), --dry-run defaults on T2, and tier gating (max s
db
Oracle Database guidance for SQL, PL/SQL, SQLcl, ORDS, administration, app development, performance, security, migrations, and agent-safe database workflows. Use when the user asks to write, edit, rewrite, review, format, debug, tune, or explain SQL; create or refactor PL/SQL; use SQLcl, Liquibase, ORDS, JDBC, node-oracledb, Python, Java, .NET, or database frameworks; troubleshoot queries, sessions, locks, waits, indexes, optimizer plans, AWR, ASH, migrations, schemas, users, roles, privileges, backup, recovery, Data Guard, RAC, multitenant, containers, monitoring, auditing, encryption, VPD, or safe agent database operations.
oracle-dba-monitoring-alerting
USE THIS SKILL WHENEVER the user mentions database monitoring, alerting, health checks, up/down or availability monitoring, listener status, alert log errors, tablespace / temp / undo / ASM / FRA space usage, archivelog generation rate, session or connection counts, blocking locks, long-running queries, datafile/tablespace offline status, invalid objects, failed logins, scheduler or cron job failures, backup success/failure monitoring, Data Guard transport/apply lag, RAC node/instance monitoring, host CPU/memory/IO, alert acknowledgment / ticket creation, custom metric or threshold tuning, OEM agent deployment, or monitoring template / metric-extension setup — EVEN IF they do not name the specific check. Also fires on ORA-19809/ORA-19804 (FRA full), ORA-00257 (archiver stuck), "MRP0 not running", "apply lag", "TNS-12541/12514 listener", "node eviction", or "tablespace 9x% full". Every self-executing check is read-only T1 (observe); threshold tuning, OEM agent deploy, metric-extension setup and ITSM ticketing
oracle-dba-multitenant-cdb-pdb
USE THIS SKILL WHENEVER the user mentions multitenant, container database / CDB, pluggable database / PDB, CDB$ROOT, PDB$SEED, open/close a PDB, PDB open mode (READ WRITE / READ ONLY / MOUNTED / RESTRICTED), SAVE STATE / DISCARD STATE, PDB clone or snapshot copy, PDB unplug / plug / manifest (.xml/.pdb), PDB relocation, CDB resource plan / inter-PDB shares, lockdown profile / PDB_LOCKDOWN, application container / application root / application PDB, PDB_PLUG_IN_VIOLATIONS, or the related views V$PDBS / CDB_PDBS / DBA_PDB_SAVED_STATES — EVEN IF they do not name the specific task. Covers PDB status monitoring, PDB open/close, save/discard state, local clone & snapshot copy, unplug-to-manifest, CDB-level Resource Manager configuration, lockdown profile rules + assignment, and the plan-only Tier-3 runbooks for PDB relocation, application-container management, plug, drop, and lockdown-profile drop. Follows the oracle-dba-common contracts: SQLcl MCP saved connection dba_ai_conn, Oracle Wallet secrets (never plaintex
oracle-dba-networking-connectivity
USE THIS SKILL WHENEVER the user mentions the Oracle listener (lsnrctl, listener.ora, listener restart/reload/start/stop, static vs dynamic registration), tnsnames.ora / sqlnet.ora / EZConnect, tnsping, service registration (ALTER SYSTEM REGISTER, service_names, local_listener/remote_listener, PMON/LREG), connectivity testing or "cannot connect to the database", Connection Manager (CMAN), SDU / SEND_BUF / RECV_BUF / network throughput tuning, or ANY of the connection errors ORA-12541 (no listener), ORA-12514 (unknown service), ORA-12505 (unknown SID), ORA-12516/12519/12520 (no handler), ORA-12170 (TNS timeout), ORA-12537/12547 (lost contact), ORA-12560 (protocol adapter), or TNS-03505 (cannot resolve name) — EVEN IF they do not name the specific task. Covers listener lifecycle, listener.ora + tnsnames.ora maintenance, static/dynamic service registration, layered ORA-12xxx diagnosis, and the plan-only Tier-3 runbooks for CMAN setup and network performance/SDU/buffer tuning. Follows the oracle-dba-common contra
oracle-dba-patching-upgrades
USE THIS SKILL WHENEVER the user mentions patching or upgrading Oracle: Release Updates (RU/RUR), one-off / interim patches, OPatch / opatchauto / datapatch, OJVM patches, patch conflicts / merge / napply / superset patches, patch rollback, Grid Infrastructure (GI) patching, rolling RAC patching, Data Guard standby-first patching, OS / host patch and reboot coordination, DST / time-zone file upgrades, database version upgrades (19c -> 23ai), AutoUpgrade (autoupgrade.jar analyze/fixups/deploy), DBUA / manual catupgrd, Fleet Patching & Provisioning (FPP / rhpctl), or cross-version / out-of-place migration — EVEN IF they do not name the specific tool. Covers read-only patch/upgrade observability and readiness reports, OPatch prereq/conflict analysis, datapatch SQL reconcile (+ rollback note), DST PREPARE analysis, AutoUpgrade analyze/fixups, OS-patch quiesce/resume coordination, and the plan-only Tier-3 runbooks for every destructive apply/upgrade/migration. Follows the oracle-dba-common contracts: SQLcl MCP sav
oracle-dba-performance-tuning-diagnostics
USE THIS SKILL WHENEVER the user mentions database is slow / hung / high CPU, performance tuning, SQL tuning, execution plan / explain plan, AWR / ADDM / ASH, SQL Tuning Advisor, SQL profile, SQL plan baseline (SPM), SPA / SQL Performance Analyzer, optimizer statistics / DBMS_STATS / histograms / data skew, top SQL / resource consumers, blocking locks / runaway / kill session, wait event / wait class analysis, latch / mutex / "cursor: pin S" / "cache buffers chains" contention, SGA / PGA / memory tuning, optimizer parameters (optimizer_*, cursor_sharing, OFE), parallel execution / PX downgrades, real-time SQL monitoring, Database Replay (RAT / workload capture & replay), partitioning strategy, or any of the symptoms/errors ORA-04031 (shared pool), ORA-04030 (PGA), ORA-00060 (deadlock), ORA-01555 (snapshot too old), "log file sync", "db file sequential/scattered read", "library cache" / "row cache" lock — EVEN IF they do not name a specific task or tool. Covers L1 triage, L1/L2 read-only diagnostics (AWR/ADDM/
oracle-dba-provisioning-installation-configuration
USE THIS SKILL WHENEVER the user mentions provisioning or installing Oracle, an RDBMS or Grid Infrastructure (GI) binary install, runInstaller / gridSetup / root.sh, ASM disk groups / CREATE DISKGROUP / +ASM, creating a database (DBCA silent/template or custom CREATE DATABASE), multitenant CDB creation, PDB creation / CREATE PLUGGABLE DATABASE, listeners / listener.ora / lsnrctl, tnsnames.ora / sqlnet.ora / tnsping, init or spfile parameters / ALTER SYSTEM SET / pfile, a parameter baseline or fleet standardization, character set / NLS / AL32UTF8, database links / CREATE DATABASE LINK, directory objects / CREATE DIRECTORY, default or temp tablespace configuration, or block size / storage standards — EVEN IF they do not name the exact task. Covers Oracle 19c and 23ai, EE, single_instance and Data Guard, on-prem. Follows the oracle-dba-common contracts: SQLcl MCP saved connection dba_ai_conn, Oracle Wallet secrets (never plaintext, never SYS/SYSTEM unless a runbook needs SYSDBA), --dry-run defaults on T2, and ti
oracle-dba-rac-high-availability
USE THIS SKILL WHENEVER the user mentions RAC, Real Application Clusters, Grid Infrastructure / GI, Clusterware, CRS, OHASD, crsctl / srvctl / olsnodes / cluvfy, OCR (Oracle Cluster Registry) or voting disk, SCAN / SCAN listener, node VIP, cluster interconnect / private network / Cache Fusion / GCS / GES, node eviction or split-brain or reboot/fence, instance start/stop on a cluster, cluster-managed database SERVICES, TAF / Transparent Application Failover, Application Continuity (AC) / FAN / ONS / FCF, connection or runtime load balancing (CLB / RLB), policy-managed vs admin-managed databases, or ANY of the codes CRS-xxxx, ORA-29740 (eviction/reconfiguration), ORA-481/ORA-29702 (CSS/clusterware), ORA-12545/ORA-12516 (SCAN/listener), or "gc blocks lost" / high "gc cr/current block receive time" — EVEN IF they do not name a specific task or tool. Covers read-only cluster/interconnect/service/eviction diagnostics, OCR manual backup, RAC instance and service start/stop/relocate, VIP & SCAN runtime ops, SCAN port
oracle-dba-routine-maintenance
USE THIS SKILL WHENEVER the user mentions routine / housekeeping DBA maintenance: recompiling INVALID objects (utlrp / UTL_RECOMP), purging old trace / alert / incident files (ADR / adrci), the standard daily/weekly health-check report, sequence cache tuning (row-cache / SQ enqueue / "sequence" waits), index rebuild / coalesce, partition add / drop / split / merge maintenance, materialized-view refresh, PURGE RECYCLEBIN, AWR snapshot / retention management, optimizer stats-history purge, or Unified/standard audit-trail purge (DBMS_AUDIT_MGMT) — EVEN IF they do not name the specific task. Covers Oracle 19c and 23ai, EE, single-instance and Data Guard, on-prem. Follows the oracle-dba-common contracts: in-DB work via the SQLcl MCP saved connection dba_ai_conn, Oracle Wallet secrets (never plaintext, never SYS/SYSTEM), --dry-run default on every T2 script (explicit --execute to act), structured logging, and tier gating (max self-execute = T2; structural DROP/SPLIT/MERGE/EXCHANGE and big segment moves are T3 runbo
oracle-dba-security-encryption-auditing
USE THIS SKILL WHENEVER the user mentions Oracle security, auditing, encryption, TDE / Transparent Data Encryption, wallet / keystore / TDE key rotation, ADMINISTER KEY MANAGEMENT, Unified or Standard/traditional auditing, audit policies, AUDIT/NOAUDIT, audit trail / AUD$ / UNIFIED_AUDIT_TRAIL / audit purge, failed-login or security-event review, CIS / STIG / security baseline / hardening, default passwords, privilege analysis (DBMS_PRIVILEGE_CAPTURE) / least privilege / unused privileges, sensitive-data discovery / PII, Data Redaction, Virtual Private Database (VPD / DBMS_RLS / row-level security), Database Vault, Audit Vault & Database Firewall (AVDF), network encryption (TLS / SSL / native / sqlnet.ora), security CVE / Critical Patch Update / datapatch, or errors like ORA-28365 (wallet not open), ORA-46630 / ORA-46658 (keystore), ORA-01017 (bad credential) — EVEN IF they do not name the specific task. Covers audit reporting, security-event review, Unified Audit configuration, bounded audit-trail purge, CIS
oracle-dba-storage
USE THIS SKILL for ANY Oracle space & storage request — tablespaces (create, add datafile, resize, bigfile), temp/undo sizing & retention, segment/table reorganization, tablespace defragmentation, ASM disk add/drop/rebalance, archive log / FRA / diag-trace / listener-log space reclamation, and segment/datafile shrink or storage-tiering (ILM/ADO) design. Invoke whenever the user mentions "out of space", ORA-01653/ORA-01654/ORA-01688/ORA-30036/ORA-39171, "tablespace full", "FRA full", "archive destination full", "shrink", "reclaim", "defrag", "rebalance", or capacity planning for storage. Follows oracle-dba-common contracts: SQLcl saved connection dba_ai_conn (wallet secrets — never plaintext), --dry-run default for all T2+ actions (--execute to act), full structured logging, and tier gating (T3 is runbook-only + human approval token). When in doubt, runbook.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.