diff --git a/src/keboola_agent_cli/changelog.py b/src/keboola_agent_cli/changelog.py
index cc6f48ff..0ab829af 100644
--- a/src/keboola_agent_cli/changelog.py
+++ b/src/keboola_agent_cli/changelog.py
@@ -27,6 +27,20 @@
         "New: config set-folder -- sugar over set-metadata for KBC.configuration.folderName; organises configs into named folder groups visible in the Keboola UI (FIIA P1-3)",
         "New: workspace list --orphaned -- lists workspaces backed by keboola.sandboxes whose sandbox config no longer exists (FIIA P1-4)",
         "New: workspace gc [--dry-run] [--yes] -- deletes all orphaned workspaces; dry-run previews without touching anything; --yes skips interactive confirmation (FIIA P1-4)",
+        "New: `kbagent storage describe-bucket` -- set KBC.description on a bucket via metadata POST (upsert-by-key, provider=user); surfaces on bucket-detail",
+        "New: `kbagent storage describe-table` -- set KBC.description on a table; surfaces on table-detail",
+        "New: `kbagent storage describe-column` -- set per-column descriptions using the `KBC.column.{name}.description` convention in table metadata; readable via table-detail column_details[].description",
+        "New: `kbagent storage describe-batch --from-file YAML` -- apply bucket/table/column descriptions in one shot; failures collected without aborting the remaining items (progress spinner in human mode)",
+        "Fix: `storage bucket-detail` now returns `description` and `metadata` fields extracted from the bucket metadata array; KBC.description (provider=user) wins over the native creation-time description field",
+        "Fix: `storage table-detail` now returns `description` and `metadata` fields extracted from the table metadata array",
+        "New: `kbagent job run --wait` uses an exponential polling curve (2s x 30 -> 5s x 48 -> 15s) instead of a fixed 1s interval (FIIA P0-3) -- matches the cadence used by the keboola-as-code Go CLI; pass `--poll-strategy fixed` to keep the legacy 1s behaviour for tests or very short jobs",
+        "New: `--log-tail-lines N` on `job run` -- on FAILED / WARNING / TERMINATED jobs, kbagent fetches the last N Storage Events (/v2/storage/events?runId=...) and surfaces them as `logTail` in --json output or `details.logTail` on errors; default 200, max 5000, 0 disables",
+        "New: `--timeout` on `job run --wait` now auto-cancels the remote job -- when the local deadline expires kbagent issues `kill_job` against the Queue and exits 7 (`JOB_TIMEOUT_TERMINATED`) with the cancelled job + logTail in error details; distinct from exit 4 (`QUEUE_JOB_TIMEOUT`) which signals the local kill ALSO failed and the remote may still be running",
+        "Client: new `fetch_job_events(run_id, limit)` wraps Storage Events API; runId resolved from the job dict (Queue v2 jobs have runId == id)",
+        "Error envelope: `KeboolaApiError` gained an optional `details: dict` payload; JSON output includes `error.details` only when non-empty so callers consume structured context without parsing the human message",
+        "Refactor: `ErrorCode(StrEnum)` in errors.py -- 49 typed constants (46 original + 3 new: JOB_TIMEOUT_TERMINATED, INVALID_FLOW_DAG, SCHEDULE_DELETE_FAILED). Every `KeboolaApiError` / `formatter.error(...)` raise site migrated from string literals to `ErrorCode.<MEMBER>`. Wire format unchanged (StrEnum subclasses str)",
+        "New: `docs/error-codes.md` -- reference for all ErrorCode members with semver policy (add=minor, rename/remove=major); `scripts/check_error_codes.py` CI guard (wired into `make check`) rejects new raw literals",
+        "New: `kbagent sync init --adopt-existing` -- idempotently adopt a `.keboola/manifest.json` written by the kbc Go CLI without overwriting it; validates manifest `project_id` against the alias token and rejects mismatch with `CONFIG_ERROR` (exit 5); falls through to normal init when no manifest exists; safe to re-run (FIIA P2-2)",
     ],
     "0.21.2": [
         "Fix: `kbagent config search` now scans `rows[].configuration` in addition to the top-level configuration body (#196) -- queries like `--query '\"incremental\": false'` previously returned zero matches for row-based components (Snowflake/MySQL/BigQuery writers, DB extractors, Google Sheets) because the service only fetched `include=configuration`; match paths are now reported as `rows[N].configuration.parameters.<key>`",