security: redact leaked Feishu Base token literal before first push

The literal token value was committed in plaintext across 4 files (agent-memory notes, a stored run artifact, and a test assertion) while documenting that it had been redacted from migrated skill files -- the documentation itself failed to redact it. Replacing with placeholders/omission ahead of pushing this repo to a remote for the first time; the pre-existing commits (2a71bd5, 0cbf4b1, 14dee05) still contain the plaintext value in history. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

security: redact leaked Feishu Base token literal before first push
The literal token value was committed in plaintext across 4 files (agent-memory notes, a stored run artifact, and a test assertion) while documenting that it had been redacted from migrated skill files -- the documentation itself failed to redact it. Replacing with placeholders/omission ahead of pushing this repo to a remote for the first time; the pre-existing commits (2a71bd5, 0cbf4b1, 14dee05) still contain the plaintext value in history. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
bboonnee
1 parent 0b35896d
Showing 4 changed files with 11 additions and 7 deletions
.claude/agent-memory/coder/MEMORY.md
.claude/agent-memory/coder/feedback_hermes_skill_migration_secret_redaction.md
.graphlemon/runs/a45c0d95-48c1-4d4d-87c7-a67827e453f9/code_diff.patch
src/test_agent_profile.py
 - [Loop-limit comment task keeps re-dispatching with zero diff](task_dispatch_loop_limit_comment_repeat.md) — src/org_loop_route.py:38 comment already exists; verify and report, don't re-add
 - [deploy_health_check timeout comment fixed 2026-08-03](project_deploy_health_check_startup_timeout_comment_resolved.md) — PH5 already shipped, old "defer" reasoning is stale; check dev-plan closure before citing phase gates
 - [deploy_health_check HTTP timeout comment fixed 2026-08-03](project_deploy_health_check_http_timeout_comment_resolved.md) — separate constant from STARTUP_TIMEOUT case, third planner cycle before a Coder applied it
-- [Hermes skill migration must redact one hardcoded Base token](feedback_hermes_skill_migration_secret_redaction.md) — `MCYAbbfuEaY03ks8F45cs53qnfd` in 3 source SKILL.md files, redact before copying into skills/
+- [Hermes skill migration must redact one hardcoded Base token](feedback_hermes_skill_migration_secret_redaction.md) — value intentionally omitted here, in 3 source SKILL.md files, redact before copying into skills/
@@ -6,8 +6,9 @@ metadata:
 ---
 When migrating `~/.hermes/profiles/.../SKILL.md` content into this repo's `skills/<skill_id>/`
-(2026-08-04, 12-skill migration task), the literal token `MCYAbbfuEaY03ks8F45cs53qnfd` (labeled
-"Base"/"Base Token" — a Feishu Bitable identifier used with `lark-cli ... --as user`) is hardcoded
+(2026-08-04, 12-skill migration task), a literal Feishu Base token (labeled "Base"/"Base Token" —
+a Feishu Bitable identifier used with `lark-cli ... --as user`; value intentionally omitted here,
+see git history pre-redaction commits if the actual value is ever needed for incident response) is hardcoded
 in plaintext in at least 3 source files: `planning-requirement-expansion/SKILL.md`,
 `qa-task-validation/SKILL.md`, `code-task-development/SKILL.md`.
@@ -6,7 +6,7 @@ index cacf348..a857fcb 100644
  - [Loop-limit comment task keeps re-dispatching with zero diff](task_dispatch_loop_limit_comment_repeat.md) — src/org_loop_route.py:38 comment already exists; verify and report, don't re-add
  - [deploy_health_check timeout comment fixed 2026-08-03](project_deploy_health_check_startup_timeout_comment_resolved.md) — PH5 already shipped, old "defer" reasoning is stale; check dev-plan closure before citing phase gates
  - [deploy_health_check HTTP timeout comment fixed 2026-08-03](project_deploy_health_check_http_timeout_comment_resolved.md) — separate constant from STARTUP_TIMEOUT case, third planner cycle before a Coder applied it
-+- [Hermes skill migration must redact one hardcoded Base token](feedback_hermes_skill_migration_secret_redaction.md) — `MCYAbbfuEaY03ks8F45cs53qnfd` in 3 source SKILL.md files, redact before copying into skills/
++- [Hermes skill migration must redact one hardcoded Base token](feedback_hermes_skill_migration_secret_redaction.md) — value intentionally omitted here, in 3 source SKILL.md files, redact before copying into skills/
 diff --git a/.claude/agent-memory/coder/feedback_hermes_skill_migration_secret_redaction.md b/.claude/agent-memory/coder/feedback_hermes_skill_migration_secret_redaction.md
 new file mode 100644
 index 0000000..d39d79a
@@ -21,8 +21,9 @@ index 0000000..d39d79a
 +---
 +
 +When migrating `~/.hermes/profiles/.../SKILL.md` content into this repo's `skills/<skill_id>/`
-+(2026-08-04, 12-skill migration task), the literal token `MCYAbbfuEaY03ks8F45cs53qnfd` (labeled
-+"Base"/"Base Token" — a Feishu Bitable identifier used with `lark-cli ... --as user`) is hardcoded
++(2026-08-04, 12-skill migration task), a literal Feishu Base token (labeled "Base"/"Base Token" —
++a Feishu Bitable identifier used with `lark-cli ... --as user`; value intentionally omitted here,
++see git history pre-redaction commits if the actual value is ever needed for incident response) is hardcoded
 +in plaintext in at least 3 source files: `planning-requirement-expansion/SKILL.md`,
 +`qa-task-validation/SKILL.md`, `code-task-development/SKILL.md`.
 +
@@ -55,7 +55,9 @@ class RenderAgentDefinitionTest(unittest.TestCase):
     def test_no_leaked_feishu_secret_or_personal_path(self):
         for org_node in agent_profile.ORG_NODES:
             content = agent_profile.render_agent_definition(org_node)
-            self.assertNotIn("MCYAbbfuEaY03ks8F45cs53qnfd", content)
+            # real value intentionally not reproduced here -- see
+            # .claude/agent-memory/coder/feedback_hermes_skill_migration_secret_redaction.md
+            self.assertNotIn("<FEISHU_BASE_TOKEN_REDACTED_PLACEHOLDER>", content)
             self.assertNotIn("/Users/yibo/Projects/lemonAI/graphagent", content)
             # the original told Coder to itself run this; the migrated version is
             # allowed to *mention* the flag only as a prohibition, not an instruction