the-agency-ai/src/claude/skills/flag
src/claude/skills/flag/SKILL.md
Quick-capture observations to a queue for later follow-up or 1B1 discussion
$ install --global
skillsauthnpx skillsauth add the-agency-ai/the-agency src/claude/skills/flagInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 22, 2026, 11:26 AM129.6s1 file scanned
SKILL.md
- description:
- Quick-capture observations to a queue for later follow-up or 1B1 discussion
<!--
Flag #62/#63: allowed-tools removed. Inherits Bash(*) from
.claude/settings.json. Restricting to specific subcommand patterns at the
skill level silently blocks agents on permission prompts the agent cannot
see — see dispatch #171 for the devex incident that surfaced this trap.
-->
Flag
Zero-friction capture of observations during work. Flag something now, discuss it later. Flags are stored in the ISCP SQLite DB (outside git) — instant capture from any worktree.
Arguments
$ARGUMENTS: The message to flag, or a subcommand (list,clear,count,discuss), or--to <agent> <message>to route to a specific agent.
Instructions
Flagging an item (to self)
If $ARGUMENTS is a message (not a subcommand):
bash $CLAUDE_PROJECT_DIR/agency/tools/flag "$ARGUMENTS"
Confirm: "Flagged: [message summary]"
Flagging to a specific agent
If $ARGUMENTS starts with --to:
bash $CLAUDE_PROJECT_DIR/agency/tools/flag --to <agent-address> "message"
The agent address must be fully qualified (repo/principal/agent). The flag appears in the target agent's queue on their next flag list or iscp-check.
Listing the queue
If $ARGUMENTS is list:
bash $CLAUDE_PROJECT_DIR/agency/tools/flag list
Displays all unread flags and marks them as read (Slack-style seen behavior). Three-state lifecycle: unread → read (on list) → processed (on discuss/clear).
Counting items
If $ARGUMENTS is count:
bash $CLAUDE_PROJECT_DIR/agency/tools/flag count
Returns unread count only.
Clearing the queue
If $ARGUMENTS is clear:
bash $CLAUDE_PROJECT_DIR/agency/tools/flag clear
Marks all flags as processed.
Discussing flagged items
If $ARGUMENTS is discuss:
bash $CLAUDE_PROJECT_DIR/agency/tools/flag discuss
Outputs the queue as a numbered 1B1 agenda and marks all as processed. Then invoke /discuss with the agenda items for structured resolution.
Flag Lifecycle
capture → unread → read (on list) → processed (on discuss/clear)
Flags are DB-only — no git payload, instant capture. They're designed for quick observations during work, not formal communication (use dispatches for that).
OFFENDERS WILL BE FED TO THE — CUTE — ATTACK KITTENS!
Related Skills
the-agency-ai/src/claude/skills/worktree-sync
business
VerifiedTrustedCommunity
Sync worktree with master — merge, copy settings, run sandbox-sync, report changes
5SKILL.mdUpdated Apr 22, 2026
the-agency-ai/src/claude/skills/worktree-list
tools
VerifiedTrustedCommunity
List all git worktrees with status info (branch, clean/dirty, deps)
5SKILL.mdUpdated Apr 22, 2026
the-agency-ai/src/claude/skills/worktree-delete
tools
VerifiedTrustedCommunity
Remove a git worktree and optionally delete its branch
5SKILL.mdUpdated Apr 22, 2026
the-agency-ai/src/claude/skills/worktree-create
development
VerifiedTrustedCommunity
Create a new git worktree with dedicated branch and bootstrapped dev environment
5SKILL.mdUpdated Apr 22, 2026